In the context of global warming, the Ross Sea's increased sea ice extent and duration during the last few decades are in contrast with the decreases observed in the Arctic from passive microwave satellite data, though in the last four years the trend has changed with a sharp decline in extent that first occurred in late 2016. But the changes in sea ice thickness, volume, and production in this region is still unclear. Without knowing the changes of sea ice thickness over Ross Sea, it is hard to evaluate how sea ice cover is responding to changing climate in this region. This dissertation contains three essays and aims to use the airborne altimetry data (IcePod 2016-2017 and IceBridge 2013) and spaceborne data (ICESat 2003-2008) to retrieve sea ice freeboard and thickness over Ross Sea, Antarctica.

The first essay is to detect leads and sea surface tie shots within leads by combining DMS images and reflectivity of Airborne Topographic Mapper L1B data and to compute total freeboard and ice thickness from the 2013 NASA Operation IceBridge data in the Ross Sea. The estimated mean sea ice thickness values are found to be in the 0.48-0.99 m range. Along the N-S track, sea ice was thinner southward rather than northward of the fluxgate, resulting in two peaks of modal thickness: 0.35 m (south) and 0.7 m (north). This supports that new ice produced in coastal polynyas is transported northward by katabatic winds off the ice shelf. The second essay is to investigate the patterns of interannual variations of total freeboard (sea ice thickness) in the Ross Sea by comparing the IcePod 2016-2017 data with both the IceBridge 2013 and the ICESat 2003-2008 data by using lowest elevation method, where a percentage of the lowest elevations over a particular length scale has been used as the local sea level. We find that compared to ICESat years, sea ice near the coast was thicker, while sea ice offshore was thinner, in the more recent IceBridge / IcePod years. The result also shows that, in general, sea ice was thicker in 2017 as compared to 2013 or 2016 (mean total freeboard is 0.02 m-0.55 m higher). The third essay is to assess the effect of different segment scale and percentage scale on lowest elevation method by using the IceBridge dataset, which provides a unique dataset with the acquisition of laser altimeter measurements with coincident high spatial resolution optical images which allows for an independent sea level validation. The results show that the segment-length scales are more influential to the sea level retrievals than the percentage scales. The 1 km segment length and 0.1 % percentage scales get the best estimation overall for all tracks.

Using Landsat 8 imagery, a six-year tendency (2013-2019) of the melt percentage of the McMurdo Ice Shelf is developed based on Normalized Difference Water Index Ice. The water pixels are mapped using an index value of 0.12 and above. The summer melt season is defined as the period of October 1st – February 28th, when liquid surface melt water presents. The retrieved melt water percentages are then compared with the daily mean and daily maximum of air temperature and solar irradiance, to determine if a possible correlation existed. The air temperature data are from automated weather stations recorded at ice runways located directly on the McMurdo Ice Shelf. Solar irradiance data (280-600 nm) are from National Oceanic and Atmospheric Administration's Antarctic UV Monitoring Network.

It is found that four of the six years when analyzed individually showed no correlation between melt water percentage with either daily averaged temperature or daily averaged solar irradiance, except for two years with good correlations: R² of 0.80 (2013-2014 season) and 0.72 (2015-2016 season), at 99% significant level. The results are similar when daily maximum temperature or daily maximum solar irradiance are used for the analysis. A multiple regression for all six years combined produces a low correlation of 0.16 for both averaged and maximum daily temperatures and solar irradiance, but at 99% significant level.

Additionally an overall slight decrease tendency for each of the three variables (temperature, solar irradiance, and melt water percentage) during the cloudless days of the study period is found. A slight increasing tendency for solar irradiance, with a slight decreasing tendency for temperature is found when the entire season data is taken into account for both daily averaged and daily maximum variables.

Vegetation research in arid and semi-arid environments tends to not only be costly but also time consuming because they are either found in isolated areas and/or covering large areas. Remote sensing has been a more cost efficient method in monitoring vegetation in these areas. Accurate data is needed for these areas not only for current research of the state of the land but also for simulation models that illustrate the interactions between land surfaces and the atmosphere. A lab experiment was conducted using a vegetation component along with an ASD Spectroradiometer to acquire reflectance values to calculate the NDVI values, which was followed up by performing a computer simulated experiment of the lab experiment.

The study was to explore NDVI values at low vegetation percentages, with two objectives: (1) if NDVI values at low vegetation percentages are good indicators of the presence of vegetation and (2) if a simulation experiment will provide similar results. Our conclusion is that NDVI is not a good indicator of vegetation presence in low vegetation cover areas and computer-based simulation does not produce the same results from the laboratory experiments.

Solar radiation is the most abundant source of energy. Measuring solar irradiance from satellites is becoming important in the field of renewable energy, when planning for photovoltaic (PV) or thermal systems. Satellites can provide long-term solar irradiance data in a large field of view compared to ground observations, although ground–based equipment can acquire data with much higher temporal resolution. Comparison of satellite estimation of shortwave downward radiative flux at the surface (G_s) against global horizontal irradiance from the ground (G_g) is important and necessary if satellite estimates are going to be used for applications in solar mapping and electricity grid integration. The amount of solar radiation reaching the earth's surface varies greatly because of changing atmospheric conditions. Clouds play a major role in regulating the amount of solar irradiance reaching the Earth's surface. It is essential to determine how solar irradiance varies with different cloud types and heights. Mapping solar energy is useful for adding solar power into power grid portfolios and the placement of solar PV system.

This dissertation consists of three interrelated studies: (1) a validation of Gs from Geostationary Operational Environmental Satellite (GOES) Surface and Insolation Products (GSIP) using G_g from two ground stations, one in the main campus of the University of Texas at San Antonio (UTSA) and the other in the Alamo Solar Farm of San Antonio (ASF); (2) a study on how solar variability relates to different cloud types and heights from GOES GSIP in the San Antonio area; and (3) an investigation the of the spatial variability of cloudiness and solar energy climatological conditions over Texas and surrounding regions based on GOES GSIP.

Satellite-derived G_s is found to be significantly (p-value<0.05) correlated with G_g at the two San Antonio locations under all sky conditions, clear-sky conditions, and cloud-sky conditions. Correlations under different cloud types (partly, water, mixed, glaciated, cirrus and multilayered) and different cloud layers (low, mid and high) are generally greater than 0.60. The most frequent clouds found in the San Antonio area are water clouds, followed by cirrus clouds. The overall good agreement of the satellite estimates with the ground observations underscores the usefulness of the GOES surface solar irradiance estimates for solar energy studies in the San Antonio area. It enables us to better understand features of different types and layers of clouds and the time when clouds are likely to occur locally in San Antonio.

In the second part, a new variability index (VI_new) having a better physical base than the one used in previous studies (VI) is introduced. The VInew performs better at all different time intervals than VI. The mean clear sky index (CSI) and solar variability are found to be correlated with the cloud types and layers with different magnitudes.

The third part demonstrates that the spatial distribution of clouds around San Antonio and beyond shows regional differences in the frequency of cloud-type and cloud-layer occurrence. The highest monthly solar energy values derived from satellite G_s are in the 151-247 kWhm^-2 range in July and the lowest in the 43-145 kWhm^-2 range in December over the study area (Texas and surrounding areas). The highest seasonal solar potential is found in summer around 457-706 kWhm^-2 and the lowest 167-481 kWhm^-2 in winter. The annual solar energy potential is 1295-2324 kWhm^-2. The solar potential is higher over the ocean than over the land. These findings are useful for the placement of photovoltaic systems and the energy management of smart grids.

Utilizing remote sensing data, specifically from the Visible Infrared Imagery Radiometric Suite (VIIRS), gas flaring can be mapped accurately within the Eagle Ford Shale (EFS) from 2012 to 2016. Utilizing this data and data provided by the Railroad Commission of Texas, the locations and numbers of gas flare are mapped and studied. Given the negative impacts that gas flaring has on the environment, the areas with high flare concentration are important to record. Temperature of flares mapped from the VIIRS is used as a proxy for understanding the specific toxic chemicals being released into the atmosphere, as temperature is a driver of various combustion reaction possibilities for natural gas. ArcGIS is mostly used for mapping and illustration, while MATLAB, Excel, and Python for calculations. It is found that the EFS had 1,704 gas flares in 2012-2015 and (reduced to) 610 gas flares in 2016, with an average temperature ranging from 1750-1900 K. This indicates that most flares were burning off CO, CO₂, NO, NO₂, and SO₂, having serious environmental and health implications. The temperatures of these flares fluctuate slightly throughout the five years, however on average they stay similar. Area change of the gas flares fluctuates throughout the years, with the largest positive change from 2015 to 2016, indicating slightly faster burning off because it was no longer economical to produce. It is found that counties McMullen and Karnes, having a high rate of production, high number of gas flare, high flare temperatures, are impacted the most by gas flaring.

The purpose of the study is to assess the rapid changes of sea ice in Arctic in recent years under a climate changing scenario, and discuss association climatic and environmental processes, based on data and information obtained with in-situ (field) investigations and remote sensing techniques. The changing of Arctic sea ice has a significant potential in affecting local and regional climate, environment, and the global ocean circulation. There are several processes associated with the Arctic sea ice, like large scale fall/winter forming and summer melting of sea ice, insulation of oceanic/atmospheric heat exchange, and various related dynamic and thermodynamic processes, like wind forcing, ocean current forcing albedo feedback, and thermohaline circulation.

To assess and understand these processes, monitoring of several large-scale physical characteristics, such as sea ice concentration (fraction of area covered by sea ice in a unit area), sea ice extent (areas that covered by sea ice with sea ice concentration higher than a certain threshold value), sea ice albedo (the fraction between energy in reflected electromagnetic wave and incident solar energy), and sea ice thickness, is crucial. With the implementation of satellite remote sensing technology, large scale monitoring of sea ice in Arctic region became available, revealing a rapid decline in Arctic sea ice, which would mean an ice-free Arctic in 50 years (with the projection of climate model) and an ongoing changing in climate in Arctic and in the future.

Other than the background introduction of chapter one and summaries of chapter six, this study discusses the monitoring and associated processes with four chapters. The first of them is focused on the areal changes of sea ice, by using in-situ and passive microwave remote sensing to monitor the sea ice extent dynamic during 1979-2012; the second is focused on in-situ measurement of the Arctic sea ice albedo in Pacific sector of Arctic in summer; the third assesse differences three different retrackers in Cryosat-2 sea ice freeboard estimation by comparison with NASA Operation IceBridge airborne altimeter data; and the fourth of them is focused on Cryosat-2 altimetry for the estimation of sea ice thickness changes in Arctic during the 2011-2014.. The results show a rapid declining Arctic sea ice extent and frequent occasions of recorded minimum extent in recent years, the ice edge is retreating by 0.022 - 0.056 degree latitude per year in average, or a 6.26-10.00 km areal decline per year, and the trend is statistically significant (0.05 significance level). The sea ice albedo change in Arctic is rapidly fluctuating and significantly affected by weather events, which can be reduced from ~0.75 - 0.80 (high albedo new snow from snowfall) to ~0.60 (surface melting due to rainfall) within one day. The sea ice thickness shows a slightly declining in 2012 and 2013 and correspond to the recorded low minimum sea ice extent in these years.

These results show an Arctic of rapid sea ice changes, with a long-term declining trend, in both extent and thickness that corresponded with each other. It is most likely that the sea ice in Arctic is in a positive feedback, that the decline in sea ice makes it much more prone to climate changes.

The aim of this work was to test a quantitative method for the evaluation of a geodiversity index, often elaborated on the basis of qualitative factors. In particular, the final purpose was to obtain an evaluation of geomorphodiversity, that is the aspect of geodiversity related to the geomorphological variety of landforms, using GIS tools. The use of georeferenced digital data, allowed the extrapolation of the morphometric data directly from a trusted source, removing subjectivity associated with analogical data. The approach used, proposed by Melelli et al (2016) considers morphometric parameters as the most important for the analysis; the only parameter that is not strictly morphometric is the geological one, but it is taken into account because the soil and rock characteristics affect the response of the land to the morphogenetic factors, controlling the resulting landforms. To compute the final analysis, which is an algebraic sum of grids, all of them were considered to have the same weight; moreover, all the parameters were converted in a variety grid, using a Focal Function tool in ArcGIS, so they can already be a measure of the diversity of the specific factor. Considering that the analysis does not imply geomorphological surveys, the final results were validated comparing the resulting data mainly with field observations. What is clear is the high spatial correlation between the geodiversity index values and the locations of the most interesting sites in the park, that have a great importance in terms of touristic value. This seems to confirm the validity of the method, that returns solid result both in the original study area in Italy, where it was developed, and in Texas, where this analysis took place.

The important role that snow cover plays in the hydrological cycle has drawn a lot of concerns from environmental researchers, especially in cold, arid, and high altitude areas, where snow melt is the primary source of water. A slight change of snow cover in high mountain area can result in great chain-reactive influence on water availability and sustainability in low altitude areas and drainage basins. Temperature and precipitation are the two defining factors that control the distribution of snow cover and its change. Therefore, how these two parameters impact the snow cover and its distribution is of great important. On the other hand, as an important indicator in global climate change, vegetation plays a linkage role in bridging land surface and the atmosphere. Predicting vegetation response to precipitation and temperature anomalies, particularly during droughts, is of great importance in semi-arid regions, because ecosystem and hydrologic processes depend on vegetation conditions. In the background of global CO₂ concentration increasing, CO₂ fertilization effect may result in non-negligible interfere on vegetation cover dynamics other than temperature and precipitation.

In the dissertation, I tested two hypotheses: (1) temperature and precipitation contributions on snow cover distribution are altitude dependent in cold and arid region and (2) CO₂ fertilization effect is increasing in arid environments.

My research focused on Heihe River basin areas: Upper Heihe River basin and Lower Heihe River basin. Data used include satellites data and ground station data, such as MODIS snow cover, land surface temperature, and normalized difference vegetation index (NDVI), shuttle radar topography mission digital elevation models (SRTM DEM), ground measured snow depth, air temperature, and precipitation. Multivariate regression analysis and statistical analysis are used to examine the relationships between climate factors, snow cover, vegetation index, and elevation and their significance levels.

In the Upper Heihe River basin, there is an altitude dependence between climate factors and snow cover area (SCA). It is found that altitude of 3800m defines the climate factor contributions. In area below this altitude, the snow accumulation is mainly controlled by precipitation (positively) while snow melting is controlled by both temperature (negatively) and precipitation (negatively); in area above this altitude, the snow accumulation is not highly related with climate factors, but snow melting is related with both temperature (negatively) and precipitation (positively).

In arid regions, C3 vegetation is assumed to be more sensitive to precipitation and CO₂ fertilization than C4 vegetation. In the second part of this dissertation, normalized difference vegetation index (NDVI) is used to examine vegetation growth in the arid Lower Heihe River Basin (LHRB), northwestern China, for the past 3 decades. The results indicate that maximum NDVI (MNDVI) of the area increases over the years and is significantly correlated with precipitation (R=0.47 and p < 0.01), not temperature (R= -0.04). The upper limit of C3 vegetation cover of the area shows a yearly rising trend of 0.6% or an overall increase of 9% over the period of 25 years, primarily due to the CO₂ fertilization effect (CO₂ rising 14%) over the same period. C3 dominant areas can be potentially distinguished by both MNDVI asynchronous seasonality and a significant relation between MNDVI and cumulative precipitation. This study provides a potential tool of identifying C3 vegetation from C4 vegetation and confirms the CO₂ fertilization effect in this arid region.

Ocean color analysis is more challenging for coastal regions than the global ocean due the effects of optical brightness, shallow and turbid water, higher phytoplankton growth rates, and the complex geometry of coastal bays and estuaries. Also, one of the key atmospheric correction assumptions (zero water leaving radiance in the near infrared) is not valid for these complex conditions. This makes it difficult to estimate the spectral radiance noise caused by atmospheric aerosols, which can vary rapidly with time and space.

This project evaluated using Landsat-7 ETM+ observations over a set of coastal bays, and allowing atmospheric correction calculations to vary with time and location as much as practical. Precise satellite orbit data was combined with operational weather and climate data to create interpolated arrays of viewing angles and atmospheric profiles. These arrays varied with time and location, allowing separate calculation of the Rayleigh and aerosol radiances for all pixels. The resulting normalized water-leaving radiance values were compared with chlorophyll fluorescence measurements made at five in-situ stations inside a set of Texas coastal bays: the Mission-Aransas National Estuarine Research Reserve.

Curve-fitting analysis showed it was possible to estimate chlorophyll-a surface area concentrations by using ETM+ water-leaving radiance values and a third-order polynomial equation. Two pairs of ETM+ bands were identified as inputs (Bands 1 and 3, and the Log10 values of Bands 3 and 4), both achieving R² of 0.69. Additional research efforts were recommended to obtain additional data, identify better curve fitting equations, and potentially extend the radiative transfer model into the water column.

With the Aquarius mission having completed its second full year of acquiring global sea surface salinity (SSS) measurements, many corrections were accounted for and biases were removed. However, some biases remain, keeping the mission from achieving its goal of +/- 0.2 psu accuracy for monthly products (150 km pixel size). Uncertainties in the Southern Ocean (among other biases) not only keep the mission from attaining such accuracy globally, but it also forces continued reliance on in situ point data sources. A Python script package is developed to process the Level 2 data for use, allowing users to target specific variables and to prepare ship and buoy data for analysis with the Aquarius data. To test the application of the scripting package, multiple assessments are completed. (1) The relationship between Aquarius brightness temperatures (Tb) and the percentage of ice and land cover is analyzed. Exponential and linear increases in Tb are observed with increasing ice and land, respectively. Little to no effect on Tb is found when there is less than 1% ice or land cover. (2) In situ SSS, in situ sea surface temperature (SST), and Aquarius Tb within a Response Surface Model are used to generate an equation to predict SSS using only Tb and SST as inputs. SSS is found strongly relying on SST, nearly removing the need for Aquarius Tb. While this does not assist in converting Aquarius Tb into SSS, the use of SST alone proved a significantly more accurate method in predicting SSS over current Aquarius estimations for the Southern Ocean. This is not to say that SST should be used to predict SSS, but rather that the two are highly linked. Discrepancies in the relationships between SSS, SST, and Tb require further investigation before SSS can be estimated from Tb in the Southern Ocean.

Using 13 years of data, from 2000 to 2012, this study examines a most complete scenario of the Italian condition about tornadoes and compares the results with previous work of Giaiotti (2006) and tornadoes in North American, the most complete in terms of knowledge and dataset allowing an interesting development of climatic framework. It has been possible to spot an increasing in number of phenomenon which can be justified in many ways: on one hand we have the growing interest in these phenomenon which certainly influenced the number of detected phenomena; on the other hand the American scenario provides a reliable hint that this growth is the result of a probable climate changing that is happening in Europe as well. In fact the significant increasing in number and intensity confirms the effective changing of boundary conditions. However it is important to underline how the awareness about this subject in Italy and Europe is still underestimated. The forecasting and detection system in real time is nearly inexistent and it should be enhanced to favor the acquaintance of this phenomenon which is currently pretty unknown in Italy and Europe.

This dissertation aims to quantify the uncertainty of remote sensing precipitation estimates. The focus is on radar rainfall estimation and satellite snow cover classification. The first part of this dissertation introduces a new approach to study the spatial, temporal and vertical variability of radar-estimated rainfall using a vertically pointing radar (Micro Rain Radar or MRR) in conjunction with a ground sensor (rain gauge) and rainfall estimates from the nearest weather radar (Next Generation Radar or NEXRAD). The MRR's direct rainfall estimates using the Mie theory has similar values when compared to the collocated gauge rainfall observations. It was found that MRR estimates are sensitive to the height resolution (the size of the vertical radar bin) and that the MRR rainfall may be biased in presence of bright band or other artifacts at higher elevations (above 2100 m).

In the second part of the research multiple radar bin integrations are used to investigate reflectivity-derived rainfall accuracy and errors from two NEXRAD radars that cover the same network of 50 gauges in the Upper Guadalupe River Basin. It is found that, in addition to the size of the integration bin, there are other sources of uncertainty such as distance from the radar, amount of rainfall, and type of the rainfall event.

The third research is a validation of a new NEXRAD rainfall product called Digital Storm Total Precipitation (DSP) using a dense gauge network in the Hill Country of Texas. The DSP is a product of high temporal and spatial resolutions intended for flash flood forecasting and warning. The validation process is based on three years of rainfall data, using statistical and analytical parameters. The accuracy of DSP is found to be highly dependent of the radar range and is also affected by seasonality, with more accurate measurements in warm season than in cold season. The DSP probability of rainfall detection is found to be always higher than gauges.

Finally, the fourth part investigates the daily snow cover product of the Ice Mapping System (IMS) at a nominal resolution of 4 x 4 km. The product's accuracy and robustness are compared against snow depth measurements from a network of 197 meteorological stations in the Colorado Plateau and MODIS satellite estimates. IMS accuracy is found to be similar to MODIS accuracy with slightly lower values during ablation and accumulation periods. IMS classification errors are also significantly comparable to MODIS errors (both at 500 m or 4 km resolutions) with the exception of unstable periods (accumulation and ablation) where IMS errors can be close to 10% higher than MODIS errors.

This study gives particular attention to spatial inequity in terms of park service. The methodology presented is a step by step approach using Geographic Information System and Spatial Cluster Analysis that are easy to adopt by any public authority. A case study of the spatial distribution of parks service in Bexar County, Texas is presented here. In addition to the traditional methods such as Buffer method and Thiessen Polygon method, this study uses Park Congestion Index and Growth Index as two indicators to locate potential future park sites. The Park Congestion Index is calculated based on the park service population using 2010 census block group population data, and the unit of Park Congestion Index is in park acres per 1,000 residents. Park Growth Index is calculated using annual park service population change rate between 2000 to 2010.

To compare the results calculated by zip code, census tract, census block group, and census block level population data of 2010, Pearson's Correlation Coefficient is used to determine correlation between any two levels. Zip code level result is obviously different from the other three levels with relatively low correlation coefficients (r < 0.70). Tract, block group, and block level results are highly correlated with each other (r > 0.98).

Spatial autocorrelation tests are performed on the reciprocal of Park Congestion Index based on two separate methods: Getis-Ord Gi* statistic and Moran's I statistic. Only hot spots (spatial clusters with high values) is observed on the output map of Getis-Ord Gi* statistic. The area of hot spots in the output map of Local Moran's I statistic, though smaller, is very similar to the areas of Getis-Ord Gi* hot spots. Significant cold spots (spatial cluster with low values) are also observed using the Local Moran's I statistic.

Raster Analysis is chosen to identify Thiessen Polygon areas with both high growth and high congestion. To avoid overlapping construction, buffer areas with walkable distance (400 m) around the perimeter of existing parks are removed from the Thiessen Polygon areas with both high growth and high congestion to locate future locations of park. Current land use is overlapped with those areas to determine which locations would be the best choice. The results indicate there are some Thiessen Polygons located in the west forest areas, where the park congestion and annual population change rate are both high. These are the candidate future park sites that public decision makers need to focus on.

This study utilizes crime rate, crime density and Location Quotient to Crimes (LQCs) in mapping the most dominant crime types across different census tracts in San Antonio, Texas. In mapping dominant crime types across different census tracts in San Antonio, crime rate and density, as well as, the Location Quotient to Crimes (LQCs) are the relevant indices used in this study; however, limitations still exist when crime rate is used as an index in mapping crimes. This is evident in this work. There are also shortcomings when crime density and the LQCs are adopted in mapping crimes. Multivariable linear regression models are used to compare the strengths of the three different indices used to analyze crime in the study. The effectiveness of crime rate is confirmed together with the association with spatial disparities of crime rates across different census tracts. Variables such as vacant housing, rental homes, unemployment and female-headed households are associated highly with crime rates across census tracts. Both violent and property crimes are closely related to vacant homes and the number of rental homes in San Antonio. This is reasonable taking into account the high rate of vacant homes and high rental market today.

Hydrological modeling has a long history in solving various surface water problems such as flood management and water conservation. In relation to water resources management, the variations of climate and physiography of Texas result in different regions with unique characteristics in surface water and groundwater resources. Hydrologic modeling plays a major role in assessing these characteristics through balancing the water budget that defines the water excess (floods) and water scarcity for water supply and irrigation. Also, water quality is crucial to the fresh water budget, thus studying water quality distribution is another point of interest to evaluate water resources. The interaction between surface water and groundwater as part of the hydrologic cycle can also be assessed by hydrologic models in term of watershed contribution to the groundwater recharge.

This study contributed to physical modeling of hydrologic processes in Texas in different ways. First, vigorous validation of physically-based distributed-parameter modeling over watersheds of different sizes was performed. Second, detailed comparison of this approach against conceptual modeling was performed to understand the advantages and disadvantages of each approach over a range of rainfall events and watershed characteristics. Third, a physically-based distributed parameter was used to improve the hydrologic simulation component of a widely accepted water quality model in simulating nutrient transport over the entire San Antonio River basin. Fourth, the physically-based distributed-parameter model was used to produce very high resolution (in time and space) estimates of recharge for the Edwards aquifer. Different vi hydrologic models are used in this dissertation to study different topics that relate to surface water and other hydrologic cycle components in different watersheds in Texas through different events and their different causes and effects in various watersheds. Some of them are semi distributed and lumped models such as Soil and Water Assessment Tool (SWAT), Hydrologic Modeling System (HEC-HMS) and physically based distributed model Girded Surface-Subsurface Hydrologic Assessment GSSHA taking the advances of GIS, NEXRAD product, remote sensing and other products such as gridded land use and soil map to achieve the highest accuracy of these models. Two different models: SWAT and GSSHA are used for water quality assessment in San Antonio River basin because the rainfall runoff simulation is needed to derive the surface water quality process, especially along the streams. The accuracy of the predictions of hydrologic models has significantly improved. The demand to solve more complicated hydrologic, hydraulic and water quality problems are increasing due to population growth and rapid development and expansion of agricultural activities.

The Tibetan Plateau (TP) in central Asia has an average elevation of more than 4000 m asl and an area of approximately 2,500,000 km², and is now called "the Third Pole" of the Earth and the "Asian water tower" with the largest ice mass outside the north and south polar regions. The 36,800 glaciers on the TP cover an area of 49,873 km² (~2% of the total area). Snow cover dynamics over the Tibetan Plateau (TP) greatly influence water availability of several major Asian rivers such as Yellow, Yangtze, Indus, Ganges, Brahmaputra, Irrawaddy, Salween and Mekong. Glaciers and snow melting is main water source of these rivers and sustains the lives of hundreds of millions of people living downstream. There are more than 1500 lakes over the TP. As a whole, the TP has undergone warming in the past three decades, especially in winter; the temperature rise of 0.3 °C per decade is twice the global warming rate. The environmental influence of global warming on the TP is evident, including accelerating glaciers/snow melt throughout almost the entire TP, permafrost degradation, and increasing temperature extremes. The temporary increase in runoff and lake level has and will continue to result in some local area flooding and devastating grasslands and villages near the lakes in the short term. However, with the continuous and accelerated shrinkage of glaciers, the Water Tower's total water storage will decrease rapidly and will eventually result in dwindling wetlands and a more serious desertification than seen previously. Due to the TP's remoteness, high altitude, thin atmosphere, and harsh weather conditions, the quantitative lake water levels and snow cover are still poorly known in the TP. In this study, the water level of 261 lakes over the TP are examined using the Ice, Cloud, and land Elevation Satellite (ICESat) altimetry data; the characteristic of water level of Qinghai lake, the larges salt lake in China is monitored using ICESat data and gauge measurement. The reason of lake level is analyzed with meteorological data; Snow cover extent (SCE), snow cover duration (SCD), and seasonal snow coverage variability, snow cover index (SCI) in basins Cedo Caka, Selin Co, Nam Co and Yumzho Yumco during hydrological year 2000-2010 are examined using the Moderate Resolution Imaging Spectroradiometer (MODIS) data.

Chapter 2 contains ICESat altimetry data are used to provide precise lake elevations of the Tibetan Plateau (TP) during the period of 2003-2009. Among the 261 lakes examined ICESat data are available on 111 lakes: 74 lakes with ICESat footprints for 4-7 years and 37 lakes with footprints for 1-3 years. This is the first time that precise lake elevation data are provided for the 111 lakes. Those ICESat elevation data can be used as baselines for future changes in lake levels as well as for changes during the 2003-2009 period. It is found that in the 74 lakes (56 salt lakes) examined, 62 (i.e. 84%) of all lakes and 50 (i.e. 89%) of the salt lakes show tendency of lake level increase. The mean lake water level increase rate is 0.23 m/year for the 56 salt lakes and 0.27 m/year for the 50 salt lakes of water level increase. The largest lake level increase rate (0.80 m/year) found in this study is the lake Cedo Caka. The 74 lakes are grouped into four subareas based on geographical locations and change tendencies in lake levels. Three of the four subareas show increased lake levels. The mean lake level change rates for subarea I, II, III, IV, and the entire TP are 0.12, 0.26, 0.19, -0.11, and 0.21 m/year, respectively. These recent increases in lake level, particularly for a high percentage of salt lakes, supports accelerated glacier melting due to global warming as the most likely cause.

Chapter 3 contains the lake's water level and changes are examined using satellite data and gauge measurement. Results show that the mean water level rose 0.67 m from 2003 to 2009 with an increase rate of 0.11 m/year, which correlates well (r²=0.90) with gauge measurements. Envisat altimetry data show a similar change rate of 0.10 m/yr, but with a 0.52 m higher due to different referencing systems. Detailed examination of 47 ICESat tracks reveals that the lake surface has an overall northward slope (~0.001o) and that the lake level increase from 2004 to 2006 was 3 times that from 2006 to 2008, with the largest water level increase of 0.58 m from Feb 2005 to Feb 2006. Combined analyses with in situ precipitation, evaporation, and runoff measurements from 1956-2009 show that an overall decreasing trend of lake level (-0.07 m/year) correlated with an overall increasing trend (+0.03 oC/year) of temperature, with three major interannual peaks of lake level increases. The longest period of lake level increase from 2004 to 2009 could partly be due to accelerated glacier/perennial snow cover melt in the region during the recent decades. Future missions of ICESat type, with possible increased repeatability, would be an invaluable asset for continuously monitoring lake level and change worldwide, besides its primary applications to polar regions.

Chapter 4 contains Snow cover variation of four salt lakes, Cedo Caka, Selin Co, Nam Co and Yamzhog Yumco, located in the central TP, are examined using MODIS data during 2000-2010 in this study. The daily Terra-Aqua composite product is further combined with two thresholds of maximum cloud percentage of 10% and composite days of 8. The results show mean cloud cover of multi-day combination product for the four basins are decreased with less than 5.5%, while the temporal resolution is preserved with an around 3 days. Time series of snow cover extent (SCE) and seasonal changes in the four basins are examined. Snow cover variations of intra-seasonal and interannual are obvious. The HY2007 shows a rich snowfall and HY2010 poor. The summer snow cover, i.e. perennial snow, indicates a decline tendency in basin Cedo Caka, Nam Co and Yamzhog Yumco over the HY2002 to HY2010 period, and no clearly variation in Selin Co. Snow cover duration (SCD) shows Nam Co basin has great spatial distribution of SCD and the maximum day of 300-350, while Selin Co basin small extent of SCD and the maximum of less than 200 days. The SCD indicates an obviously relationship with elevation of terra. Snow cover index (SCI) shows an increase then decrease change for basin Cedo Caka in HY2007, Selin Co in HY2005, Nam Co in 2003, respectively, Yamzhog Yumco an overall increase except for HY2010.

Sea ice, most simply stated, is frozen ocean water. It covers about 10% of the Earth's surface at peak extent, varying seasonally between the Northern and Southern Hemispheres. Sea ice plays an important role in the global climate system forming an insulating cover for the polar oceans, thereby regulating heat and energy transfer, and gas exchange with the atmosphere. The highly reflective surface of sea ice and its snow cover (albedo) reduces the Earth's absorption of solar radiation, instead reflecting it back toward space. Unlike the recent documented decline in sea ice extent for the Arctic, the Antarctic has not experienced a significant decline over the 30 year period of satellite monitoring. However, regional trends in Antarctic sea ice extent indicate increases in areas like the Ross and Weddell Seas as compared to decreases in the Bellingshausen and Amundsen Seas. Because of the remote location, logistical constraints, and extreme conditions, the Antarctic sea ice zone remains one of the least visited areas of the planet. For these same reasons it is crucial to utilize satellite remote sensing for long term monitoring of the environmental conditions in Antarctica to understand the impacts of climate change.

In this research, passive microwave remote sensing products generated for the Antarctic sea ice zone from the Advance Microwave Scanning Radiometer- Earth Observing System (AMSR-E) sensor were compared with various in situ field measurements, both from previous Antarctic campaigns in the published literature and as obtained during the Sea Ice Mass Balance in the Antarctic (SIMBA) project during the International Polar Year (IPY) 2007-2008. Data gathered during the SIMBA project was used to understand the geophysical processes occurring in the sea ice and snow cover of the Bellingshausen Sea and to provide a physical basis for modeling of microwave emissions. The microwave emissions modeling is an important element in understanding how these sea ice and snow processes affect microwave signatures as retrieved from space-borne sensors. The coarse spatial resolution of passive microwave sensors necessarily involves a mixture of signals that are confounded by a large number of physical effects that alter the dielectric properties of the medium. These effects currently provide limitations on the utility of passive microwave remote sensing for monitoring the Antarctic sea ice zone.

In Chapter 2, the AMSR-E sea ice temperature product was compared with AMSR-E snow depth product and previous in situ field measurements. The comparisons were not intended to provide a strict validation of remote sensing products, but to evaluate the physical context of the remotely sensed data and examine potential trends. From examination of the data, it was found that the AMSR-E sea ice temperature product conflicted with several generally observed sea ice properties. The apparent contradictory behavior of the satellite data product is indicative of radiative temperature behavior related to changes in emissivity within the ice pack. Further comparisons of the AMSR-E sea ice temperature product with in situ temperature data from Ice Mass-balance Buoys (IMB) from two Antarctic field programs showed no correlation. However, apparent response of sea ice temperature product to snow/ice interface flooding events was noted. The study was a contributing factor in NASA's response to remove the AMSR-E sea ice temperature product from available datasets in 2008 due to "inherent ambiguities between changes in the physical temperature of the ice and changes in emissivity."

In Chapter 3, an important sea ice process related to the formation of "gap layers" within Antarctic sea ice was examined and modeled. Gap layers are horizontal voids that develop internally within the sea ice structure, often filled with decaying sea ice, saline slush, and a microbial biological community that thrives on the available nutrients. Gap layers are commonly observed in summer melt conditions in Antarctic sea ice, but are not widely observed in the Arctic. A thermodynamic model was developed based on a typical summer temperature gradient reversal in the snow pack and sea ice, typical salinity profile and heat flux to explain the internal melting of sea ice and formation of gap layers. The modeled rates of gap layer formation generally agreed with published field observations. This chapter was published in Geophysical Research Letters, Volume 35, Number 11, 2008.

In Chapter 4, an overview of the Sea Ice Mass Balance in the Antarctic (SIMBA) experiment is provided detailing various geophysical measurements and the observed snow and sea ice processes occurring during the winter-spring transition in the Bellingshausen Sea. Time series measurements were obtained for snow and sea ice conditions during a 27-day drift station through a number of atmospheric cycles of warming and cooling that are typical of the season for this region. Characteristic sites representing the range of snow and ice conditions on the drifting floe (Ice Station Belgica) were sampled at regular intervals to understand changing conditions in response to the atmospheric events. Detailed snow and ice properties and structure, including high resolution time-series records of snow and ice temperature were obtained from ice mass-balance buoys (IMBs) and other sources to record the changes. The snow and sea ice processes documented during SIMBA were compared to published literature and used to understand seasonal characteristics of the region. The time-series snow and ice data were further utilized in Chapter 5 to provide detailed modeling of passive microwave emissions. This chapter is published in Deep Sea Research II (doi:10.1016/j.dsr2.2010.10.027).

Chapter 5 presents the results of microwave emission modeling performed using the SIMBA field data, specifically processes that are commonly observed in the Antarctic sea ice zone that are considered to have an impact on passive microwave retrievals from space. The field data were processed and used as input to the multi-layer Microwave Emission Model of Layered Snowpacks (MEMLS) as modified in 2006 to incorporate the complex dielectric behavior of saline snow. In several model cases of varying snow cover thickness, the flooding of the snow-ice interface with sea water to form a saline slush layer in the snow cover was simulated. Additionally, a model case including brine wicking at the surface of first year sea ice with thin snow cover was simulated. These processes (related to Chapter 2) have been attributed to anomalous behavior in the AMSR-E sea ice temperature product and were identified as sources of error in other passive microwave sea ice products. The modeling results indicated that brightness temperature at low frequencies (6.9 and 10.7 GHz) showed a large decrease (on the order of 15 to 30°K) and are consistent with previous laboratory experiments. Further time-series examination of microwave emissions from space, cross frequency and polarization responses, has potential to indicate areas with widespread snow/ice interface flooding. However, additional research and ground based observations are required to validate the approach.

Sea ice is a fundamental component of the Earth's systems that cannot be ignored in the large scale environmental predictions of future climate conditions. Sea ice is a complex material and has major influences on global climate with its large maximum extent and seasonal change. In contrast, sea ice is also vulnerable and sensitive to global climate change. The Antarctic sea ice zone remains one of the least known regions of the Earth's surface. Both passive and active microwave remote sensing have provided useful information about sea ice properties in both Polar Regions and their trends of change over 30 years. Satellite laser and radar altimetry measurements are nascent technology and have been used less than a decade. For Antarctic sea ice, however, work on computing ice properties from satellite algorithms are still in a developmental and quasi-validated state. In this research, remote sensing validation based on comparisons with surface based data has been done for quantitative monitoring of the ice properties. Various satellite products consisting of passive microwave, active microwave, and high resolution visible imagery were used and compared with in-situ measurements collected during scientific Antarctic cruises, conducted during International Polar Year (IPY) 2007-2008. In-situ measurements were used as ground truth data to validate satellite measurements, in terms of looking at sea ice concentration, sea ice extent, and sea ice types. In addition, National Ice Center (NIC) ice edge data was used to compare and compliment satellite and in-situ measurements. In chapter 5, data sets on small-scale profiles on surface elevation gathered from ships were standardized. This data used to provide a quantifiable method for observing sea ice, from all regions of the Antarctic sea ice zone to develop relationships that test existing remote sensing algorithms, evaluate alternative algorithms and provide error estimates on sea ice thickness derived from existing algorithms.

Chapter 2 presents the comparison of ice extent/ice edge data from the NIC and the AMSR-E (Advanced Microwave Scanning Radiometer-Earth Observing System) passive microwave products using the Antarctic Sea Ice Process and Climate (ASPeCt) ship observations from the Oden expedition in December 2006 as ground truth to verify the two products during Antarctic summer. Ice edge location comparison has also been made between the two data sets, ship ice observations and NIC daily ice edge products. NIC analyses rely more heavily on high resolution satellite imagery such as active radar and visible imagery when visibility (clouds) allows. From these comparisons, a quantitative estimate of the differences in summer ice extent between the two remotely obtained products, AMSR-E and NIC ice edge, over the larger West Antarctic sea ice zone, has been obtained.

Chapter 3 evaluates the comparison of ASPeCt ship based observations (conducted during Sea Ice Mass Balance in the Antarctic (SIMBA) 2007 Antarctic cruise) with coincident satellite active and passive microwave data. We combined visual ship-based observations of sea-ice and snow properties during SIMBA with coincident active and passive microwave satellite data with the aims to a) derive typical radar backscatter ranges for observed sea-ice types and ice type mixtures, b) improve our knowledge about the radar backscatter of different ice types in the Bellingshausen Sea at early-middle spring, c) interpret AMSR-E snow depth over these ice types, and d) identify the potential of the investigated active microwave signatures for a synergy with AMSR-E data to eventually improve the snow depth retrieval.

Chapter 4 presents the validation of remote sensing measurements of ice extent and concentration with ASPeCt ship-based ice observations, conducted during the SIMBA and the Sea Ice Physics and Ecosystem eXperiment (SIPEX) International Polar Year (IPY) cruises (Sept-Oct 2007). First, the total sea ice cover around the entire continent was determined for 2007-2008 from Advanced Microwave Scanning Radiometer - Earth Observing System (AMSR-E) passive microwave and National Ice Center (NIC) charts. Second, Antarctic Sea Ice Processes and Climate (ASPeCt) ship observations from the SIMBA and SIPEX expeditions in the austral end of winter – beginning of spring 2007 are used as ground truth to verify the AMSR-E sea ice concentration product provided by both the Enhanced NASA Team Algorithm (NT2) and Bootstrap Basic Algorithm (BBA).

Chapter 5 presents supplemental analysis related to the baseline thickness of Antarctic sea ice on a circumpolar basis from field measurements. In this part, our objectives were (1) Develop statistical relationships between surface elevation (snow freeboard), ice elevation (ice freeboard) and mean sea ice thickness using previous and newly obtained Antarctic sea ice profiles and examine these relationships for any consistent regional trends, (2) Derive sea ice thickness from profile elevations, using buoyancy equation, to determine error estimates compared to measured thickness; compare error estimates between the thicknesses derived using statistical relationships (Objective 1) and buoyancy theory where the additional term for the density of the slush layer is needed, when surfaces are flooded from snow loading.

In this study, MODIS (Moderate Resolution Imaging Spectroradiometer ?Earth Observing System) land surface temperature (LST) products of eight-day composite images at 10:30 am, 1:30 pm, 10:30 pm, and 1:30 am are used to study urban heat island (UHI) phenomenon over four major Texas cities (El Paso, Dallas-Ft. Worth, Houston, and San Antonio) from the Summer of 2000 to the Fall of 2008. The eight-day LST products are used to generate spatial maps characterizing the temperature distribution and UHI spatial extent for each city.

The results indicate that (1) UHI can be observed in night time images in Dallas-Ft. Worth, Houston, and San Antonio all year long; the intensities of UHI are larger in night times of spring and summer seasons than those of fall and winter seasons; (2) UHI consistently appears in night time images in El Paso-Juarez and the intensities of UHI are similar in all seasons; (3) the day time images contain large amounts of cloud contamination in Dallas-Ft. Worth, Houston, and San Antonio which make the use of day time images to map UHIs in those cities difficult; and (4) from the temperature climatology maps (seasonal mean and yearly mean) of the 8 years, it is found that Aqua/MODIS LST products in the night time (1:30 am) is the best for mapping UHI for all four cities and that the spatial extent and pattern of the UHI differs from the expected HI usually centered in downtown of a city.

Antarctic sea ice dynamics are largely affected by ocean and wind forcing because it is surrounded by the open ocean, whereas Arctic sea ice is surrounded by a land mass. Opportunities to study the variations in sea ice conditions are infrequent due to the remote location and relative expense. For that reason, it is necessary to develop methods that will allow efficient and effective collection of sea ice measurements for integration with large-scale models and validation schemes for satellite products. The use of automated devices will improve estimates on sea ice trends for the Antarctic region.

Collecting ice thickness distribution trends from drilling transects can be a cumbersome ordeal and provides very little data over a large area. Therefore, it is necessary to consider using automated devices to assist in further data collection for future cruises. The first part of this study focused on compiling various datasets from the SIMBA cruise (Sea Ice Mass Balance in the Antarctic) which included ship-based sea ice observations, an electromagnetic induction device (EM-31), and video imagery (Evaluative Imagery Support Camera (EIS Cam 1)) to evaluate which automated device provided the best method to measure the sea ice thickness distribution. Remote sensing applications were used for image analysis with data from EIS Cam 1 to measure thickness of overturned ice that was being broken by the ship's hull. The thickness distribution of EIS Cam 1 and the EM-31 were then compared with the ASPeCt (Antarctic Sea Ice Processes and Climate) ship-based observations to evaluate how well each device performs. Since the footprints of three datasets were different from each other, only the frequency of the ice thickness distribution was gauged and compared. The EM-31 data overall performed better than the video imagery, for the reason that it was measuring ice conditions far enough from the ship's base, where it was capable of measuring ridged and deformation features not present in the video footprint. The study also shows potential good results for level ice up to 2.50m, although the ship's track will be biased toward thinner ice and may cause the EM-31 to oversample thin ice compared to the thicker ice surrounding the narrow track. However, under those conditions the EM-31 will act as an appropriate supplement for ASPeCt visual observations taken hourly from the ship's bridge.

The second part of this study evaluated sea ice concentration data recorded with the use of video imagery (EIS Cam 2) compared with ship-based ice observations. Images from the inbound and outbound transects were classified using techniques provided by Weissling et al. (2009) to ascertain the amount of error between camera measurements and ship-based observations. Analysis of these comparisons found poor correlations during evening conditions due to highlights and shadows generated by ridging, deformation features on the sea ice, and darker lighting conditions, in which EIS Cam 2 either underestimated concentration values up to 30% when the ASPeCt ice concentration was over 80% or overestimated ice concentration up to 60% when ASPeCt ice concentration was less than 80%. Large over- or under-estimation from ASPeCt observers was also possible due to the night condition, which was seasonally dependent. However, there was an overall good agreement between both datasets during the day time where EIS Cam 2 and ASPeCt differed approximately ~5% (inbound track) or 10% (outbound track). The errors with the datasets were related to the coarse resolution of ASPeCt parameters and the inability for the EIS Cam 2 to distinguish shadows (from ridges or the ship) and/or very thin ice types from open water when the unsupervised classification method was applied. However overall, EIS Cam 2 is advantageous in providing a constant record of sea ice concentration for a large field of view that can be used to support quality assurance purposes for ASPeCt records or supplement future cruises without an observer.

Frequent and long term snow observations and accurate snow cover measurements are critical for snowmelt-runoff prediction, operational flood control, water supply planning, and water resource management in basins where snowpack as a dominate water resource. Compared with conventional snow cover measurements, satellite remote sensing images are particularly well-adapted to the monitoring of snow cover over continuous spatio-temporal scales. The purpose of this study was to test simple and robust approaches for improving existing MODIS snow cover products and to develop parameters for studying spatial and temporal variability of snow cover from remote sensing images. The ultimate goals were to obtain near real time snow cover products with high spatial resolution and still accurate enough for regional applications, and to derive the maps of snow cover parameters, which can provide insights to better understanding spatial and temporal characteristics of snow cover and are suitable for climatology and hydrology modeling. Since the middle of the 1960's, a number of satellite-derived snow products from visible and infrared spectral as well as from passive microwaves have been available, with a few available in near-real time through the Internet. The quality of those products varies considerably in terms of sensors and platform characteristics, image processing procedures and snow classification techniques. Optical sensors such as AVHRR, MODIS, SPOT and Landsat TM have been well developed to produce snow cover maps with high spatial resolution. But the accuracy of snow mapping is largely affected by cloud cover and illumination conditions. Thermal infrared images can obtain images at night times but also remain sensitive to the presence of cloud cover. Passive microwave sensors, such as SMMR, SSM/I and AMSR-E can penetrate cloud and acquire images in all weather conditions but their coarse spatial resolution hinders their applications on operational hydrological modeling and snow-caused disasters monitoring. Active microwave sensor, such as RADARSAT, has fine spatial resolution but the complex interactions between the radar signal and snow pack properties strongly limit the processing and interpretation of radar images. Presently, because of the advantages and disadvantages inherent to each data type employed to monitor snow cover from space, the most favorable solution is probably in using multiday combined and/or multi-sensor combined products. This dissertation consists of four research papers but in a continuous way of improving the snow cover mapping accuracy (Chapters 3 and 4), developing parameter maps of snow cover spatial and temporal variability (Chapter 5), and their applications in the Qinghai-Tibet Plateau (chapter 6).

In Chapter 3, an advanced method is presented to combine daily Terra-Aqua MODIS and Aqua AMSR-E snow products for generating new daily cloud-free snow cover (SC) and snow water equivalent (SWE) products, both in 500 m spatial resolution. This method consists of three major processes: unifying codes, combing products, and redistributing SWE based on sub-pixel analysis. The method was tested in Fairbanks and Upper Susitna Valley, Alaska area for one water year (October 2006 to September 2007). The result confirms that MODIS has high classification accuracy in cloud-free condition and that the daily Terra-Aqua MODIS combination reduces cloud contamination by 12.2% over MYD10A1 (Aqua) and 7.2% over MOD10A1 (Terra). The evaluation against in situ observations at all weather conditions shows that the snow accuracy of the new SC products is 86%, which is much higher than the 49% of the Terra-Aqua daily combined products. The validation of current SWE products demonstrates that the accuracy of AMSR-E is 68.5% and it tends to overestimate SWE. Redistribution of SWE, based on sub-pixel analysis of AMSR-E pixels, not only generates the new product at 500 m spatial resolution, now more suitable for basin and regional monitoring and modeling, but also slightly increases the accuracy of the SWE estimations. The new daily cloud-free products (SC and SWE) are therefore significant improvements to the standard MODIS and AMSR-E snow product series. This result is submitted to Remote Sensing of Environment for publication.

In Chapter 4, various approaches including daily combination, adjacent filter, fixed day combination, multiday combination and multi-sensor (with AMSR-E) combination were tested to reduce or remove cloud obscuration. Daily combination approach, termed the combination of Terra and Aqua, merges the two MODIS snow cover products on a pixel basis, using a so-called priority approach. The pixels classified as clouds in the Terra MODIS images are updated by the Aqua MODIS pixel values if the corresponding Aqua MODIS pixel values are snow or land, and vice versa. Adjacent filter approaches, retrieve the cloud pixels using the information from the pervious and the next day. Fixed day combination approaches, replaces clouds pixels by the most recent preceding non-cloud observations at the same pixel within predefined fixed temporal windows of 2, 4, 6, 8 days. Multiday combination approaches, also used the priority principle while with a flexible temporal window that controlled by two user-defined thresholds, namely, cloud cover percentage (P) and maximum composite days (N), depending on the purpose of application. Multi-sensor combination approach uses AMSR-E snow cover product to map those areas where MODIS can not map because of clouds and polar nights. The performances of different snow cover maps generated by these methods were quantitatively evaluated through cloud and snow percentages, misclassification error, overall accuracy index against in situ observations at 244 SNOTEL stations over the northwestern United States during three hydrological years from 2006 to 2008. The average cloud coverage (52%) of Terra MODIS snow cover product is reduced to 44% for daily combination, 36% for adjacent filter, 27% for 2 days combined, 13% for 4 days combined, 7% for 6 days combined, 4% for 8 days combined, and 7% for multiday combined, while the overall accuracy (in clear sky conditions) of them is 90.4% to 89.8%, 90.1%, 89.3%, 88.8%, 88.2%, 87.7%, 89.1%, respectively. All the results indicate that the adjacent filter approach is an efficient method to reduce cloud contamination and improve image classification accuracy. And along 2 to 8 days combinations, the cloud coverage decreases, but sometimes so does the accuracy as progressively more data are merged. However the multiday combined method can keep the balance of classification accuracy and cloud reduction. The multi-sensor combination with snow product from AMSR-E makes the accuracy decrease to 79.1% for daily adjacent filter and 86.3% for multiday combined, as compared with those in clear sky conditions, and the spatial resolution also reduce because of the coarse spatial resolution of AMSR-E snow products. But this is by far the practical approach to eliminate cloud completely. Although the adjacent filter, multiday and multi-sensor combined methods were designed for application in the processing of data from MODIS and AMSR-E, it also can be used in merging other optical data such as AVHRR, Landsat TM and passive microwave data such as SSMR, SSM/I, and for future NPP and NPOESS missions. The choice of different approaches will depend on the purpose of different applications. The results in this thesis give guidance on this choice.

In Chapter 5, snow cover parameters are developed based on different products and performances are validated based on the 244 SNOTEL stations. Snow cover parameters such as Snow Cover Onset Data (SCOD), Snow Cover End Data (SCED) and Snow Cover Duration (SCD) are important factors in hydrological modeling and weather forecasting. The time series of these snow cover parameters can also give insight to the effects of global warming. Two new proposed methods and one published method were tested to derive those snow cover parameters from cloud-free AMSRE, MODIS/AMSR-E blended snow cover products (MODISMC_AE) and low cloud MODIS multiday snow cover products. Those methods are also applied and evaluated over the northwestern Untied States against the in situ observations at 244 SNOTEL stations during 2006-2008 hydrological years. Method 1 is to calculate the snow cover parameters from cloud-free AMSR-E and blended MODIS/AMSR-E snow cover products. The results indicate that the difference between AMSR-E and in situ measurements is larger than that between MODIS/AMSR-E blended snow cover products (MODISMC_AE) and in situ measurements, because of the coarse spatial resolution of AMSR-E (25 km). The no difference between AMSR-E and in situ observations for SCD, SCOD and SCED is 58%, 60% and 39%, respectively. A published method was used to detecting snow cover parameters from low-cloud MODIS multiday combined snow cover products. However the results from our study are not so good since this method has the regional limitation. Using this method, the no difference between images and in situ observations for SCD, SCOD and SCED is 41%, 33% and 14%, respectively. Method 2, based on traditional definitions of SCD, SCOD, and SCED, is to derive snow cover parameters from low-cloud MODIS multiday snow cover products. The no difference between images and in situ observations for SCD, SCOD and SCED is 72%, 68% and 49%, respectively. And the distribution of estimation errors is similar as the results from MODISMC_AE derived by Method 1, which suggest that the main source of the errors comes from the difference between image and in situ observations, and not from the method itself. The correlation analysis between errors in parameters estimation from MODISMC using the method 2 and MODISMC's under- and over-estimate errors, which are used to evaluate the difference between image and in situ observations, indicates that they have high correlation. This result further confirmed that the errors in parameters estimation are mainly from the difference between imagery and in situ observations. Moreover, because MODISMC has relatively higher spatial resolution, the results from MODISMC by method 2 have the best performance. In the derived snow cover parameters maps, mountainous areas are clearly visible. This means that snow covers early, melts later and lasts longer in the high mountains. In a word, these new snow cover maps and snow cover parameters will allow us to estimate spatial-temporal change of snow cover and snow melt more accurately, and can also apply in the operational snowmelt runoff forecasting, the calibration or validation of hydrological models and the response researches of global warming.

In Chapter 6, the developed methods are applied to Qinghai-Tibet plateau in China. The eastern Plateau is the headstream of three major Asian rivers: Yangtze River, Yellow River and Lancang (Mekong) River. Snow cover and glacier are most sensitive to climatic and environmental changes. Snow cover changes in this area have a significant impact upon Asian monsoon, droughts and floods in South Asia and East Asia. So it is important to establish such a baseline data of snow cover and glacier distribution and then their trend and changes can be studied. Utilizing the newly developed multiday combined method, "cloud free" snow cover products in 500 meter spatial resolution for the period of 2000-2008 were generated. The new product has much lower cloud percentage (5.8%) which can fulfill the needs of most snow monitoring and relative high temporal resolution (one image/2-3 days). Based on these new snow cover products, snow cover parameters such as Snow Cover Onset Date (SCOD), Snow Melting Onset Data (SMOD), Snow Cover Duration (SCD), Perennial Snow (PS) and Snow Cover Index (SCI) were derived by using the method 2. Through comparing SCOD and SCED maps of the same year and different years, the process of snow falling and snow melting can be studied. Generally, except the perennial snow and glaciers, it is found that snow covers start in September or October and end after May on high-elevation area. However in the low-elevation area, snow usually starts in November and end in February or March, while there are some areas that snow only remains less than one month. The SCD maps demonstrated that the annual SCD varied, but generally increases with elevation increasing. The major snow cover areas include the south Tanggula Mountain, the middle Bayan Kara Mountain and the eastern part of Yellow River headstream watershed. And there are some perennial snow cover/glaciers at the peak of Bayan Kara and western part of Tanggula Mountain. In addition, Yangtze River headstream watershed has the biggest SCI since it has the biggest area and the SCI of Lancang River is the smallest. From 2000 to 2008 hydrological years, the SCI in 2005 is the biggest followed by 2008. However there are no evident trends of snow change from 2001 to 2008 hydrological years. A longer time series of data is needed to study the temporal trend as well as the spatial variability of snow cover. This study, therefore, has laid a solid foundation for future work.

The growing of economy in Central Texas area resulted in the degradation of environment. The pollutants, bacteria loading and water quality assessment is required to evaluate and predict the environment quality. Precipitation is the main source of storm discharge and runoff, which becomes a critical input to several hydrological, ecological, climatic and flood prediction models. The purpose of this study was to improve the spatial resolution as well as the accuracy of the NEXRAD MPE products.

The first part of this study was to improve the resolution of the original 4km x 4km NEXRAD MPE products by the means of downscaling the radar products into 1km x 1km. The downscaling algorithm estimates precipitation distribution without prior knowledge of the atmospheric setting. It auto-searches precipitation spatial structures and atmospheric effects by incorporating elevations from a digital elevation model (DEM) into precipitation maps. The downscaled precipitation fields were examined based on different time scales: hour, day and storm period. Three downscaled precipitation fields are in good agreement with the original 4km x 4km NEXRAD precipitation fields. However, the correlation coefficients between gauge rainfall and the downscaled radar products kept the same or even less than those between the gauge rainfall and the original radar products. This does not mean that the accuracy of downscaled rainfall did not change or even less. Other approaches should be used to explore the accuracy change. However, the regression algorithm may be an efficient model in capturing the variability of spatial rainfall distribution in mountainous area, but not as efficient in flat area. Incorporating the topological information from DE M may be more effective for mountainous regions.

The second part was to improve the accuracy of NEXRAD MPE products in capturing rainfall. The major difference between this study and precious study (Wang et al, 2008) is that the validation and correction of the current study is based on the fact that the spatial and temporal continuity of precipitation is reserved. This is very important, since for hydrological modeling, the continuous and spatially distributed precipitation data is recognized as a significant input. Thus, this part aimed at conducting continuous hourly spatial and temporal evaluation of the accuracy of NEXRAD precipitation estimates and comparing 4 different interpolation methods (Bias Adjustment (BA), Simple Kriging with varying Local Means (SKlm), Kriging with External Drift (KED), and Regression Kriging (RK)) for incorporating rain gauge measurements into NEXRAD MPE products. Four evaluation parameters (Percentage Bias, Mean Absolute Error, Coefficient of Determination, and Nash-Sutcliffe efficiency) were used to evaluate the performances using the observed rain gauge data as constraint. The comparison results show that the average performance of SKlm is similar to or better than the other methods. KED is a most vulnerable method and we have to use it carefully. It is worth noting that no one method can consistently outperform the other methods in terms of all evaluation coefficients, for all time steps, and at all rain gauges. In practical application of NEXRAD precipitation products, if there is plenty of time and computational resource, it is suggested to implement multiple methods to correct the original NEXRAD data, and choose the one with best performance for some specific objectives. Otherwise, SKlm is the preferable method for incorporating rain gauge measurements into NEXRAD MPE products.

Overall, it is clear that incorporating secondary source into the original NEXRAD MPE products can improve the resolution and accuracy of original products. To satisfy the model requirements of high quality precipitation data, combination of incorporating both DEM and rain gauge measurements can be a good approach. SKlm is generally a good method in precipitation interpolation, since it is easy to implement and achieve desire results.

This study investigated the potential of Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) to estimate root zone soil moisture at native in situ measured sites, and at increasingly distant sites within the same climatic setting. In situ data were obtained from Soil Climate Analysis Network (SCAN) sites near the Texas-New Mexico border area, and NDVI and EVI products from the Moderate Resolution Imaging Spectroradiometer (MODIS) sensor on board the Terra satellite. Results show that soil moisture values at the same depth are highly correlated (r = 0.53 to 0.85) at distant sites as far as 150 Km from the native site, and NDVI and EVI are highly correlated at each site (r = 0.95 to 0.98). Raw time series has higher mean correlations than deseasonalized time series at every depth. Deseasonalized time series using NDVI and EVI are significantly correlated with soil moisture at distant sites (NDVIr = 0.35 to 0.73). The correlation reaches maximum value when the Vegetation Index (VI) lags soil moisture by 5 to 10 days. NDVI has a slightly higher correlation with soil moisture than EVI. Values of r decrease with distance from the native site. Regression analysis was also conducted using deseasonalized NDVI and deseasonalized soil moisture time series with a 5 day time lag of NDVI. The model estimated soil moisture at all depths, with adjusted R² ranging from 0.44 to 0.59. Overall, deseasonalized NDVI values produce consistent results, and show that NDVI can estimate root zone soil moisture at distant sites in the study area.

This research conjectures that streamflow in large (> 250 km2) semi-arid watersheds, in south central Texas, can be reasonably estimated with statistical regression methods based on a model solely parameterized by remote-sensing derived proxies for landscape moisture state and radar (NEXRAD)precipitation estimates. The utility of such an approach is obvious: estimating streamflow in watersheds for which no other flow records exist. The structure of the research methodology is based on three stages, 1) test the initial hypothesis that remote-sensing-derived biophysicals of the landscape (in a test watershed) are sensitive to land surface soil moisture state, and compare the resulting model against a benchmark model for streamflow estimation, 2) reassess the significance of the previous model parameters (and evaluate new parameters) in the same test watershed for an extended calibration and validation period, and 3) test for the transferability of parameters to three regionally proximate watersheds of varying dimension and environmental condition.

Chapter 2 reports on the development of the original model, including a discussion of the forward stepwise regression approach for the initial selection of model parameters. Spatially distributed NEXRAD radar precipitation estimates are introduced as an improvement over local gauged estimates. A benchmark comparison model, the Natural Resource Conservation Service curve number method, is built with multiple approaches for assessing antecedent moisture condition. The curve number method model is compared to the remote sensing model, demonstrating that the remote sensing model performs at least as well as the benchmark model, as assessed with standard efficiency criteria.

Chapter 3 continues the basic line of research but significantly expands on the evaluation of other remote sensing derived parameters potentially sensitive to landscape moisture status. The effects of deseasonalizing (removing long-term mean seasonal variation) the parameters are assessed. A final parameter set consisting of the radar precipitation estimate and two biophysical parameters derived from MODIS/TERRA satellite imagery, a land surface temperature and a vegetation moisture stress index, are regressed against a 4-year streamflow gauge record in the test watershed. The resultant calibrated streamflow estimation equation is applied to a 15 month follow-on period for model validation.

Finally, Chapter 4 evaluates the validity of applying the test watershed regression equation to three regionally proximate watersheds for the generation of a 4-year modeled flow series for each watershed. The efficiencies of these series are assessed against known gauged flow records. The results indicate that the estimation equation performed reasonably well at predicting streamflow for the watershed most similar to the test watershed in environment and dimension, with an expected loss of efficiency in the watershed for which environment and climate were most dissimilar. The estimation equation was recalibrated with parameter sets specific to each watershed with much improved results.

Chapter 5 represents a significant departure from the line of research presented in chapters 2–4, although it is also framed within the realm of remote sensing imagery analysis. An imagery acquisition system was built and deployed (by the author) for the monitoring of sea ice during the Sea Ice Mass Balance in Antarctica (SIMBA) expedition of 2007. A post processing image analysis methodology was developed to accurately quantify sea ice state, such as ice concentration, floe size, and area of deformed ice. The techniques developed represent ground-breaking research for sea ice monitoring and analysis.

The purpose of this study is to improve our understanding of Next Generation Weather Radar (NEXRAD) precipitation products by comparing them to high accuracy of rain gauge network, for the estimation of precipitation quantity and spatial rainfall distribution. The first part of the study was the calibration and installation of double-gauge platforms within a single radar cell (1 km by 1°), with some preliminary data analysis based on collected data. The second part was a comparison study of a network of 50 rain gauges with NEXRAD's Digital Storm Total Precipitation (DSP) product.

The first part consists of two phases: a testing phase for hardware calibration and development of data processing based on one double-gauge platform on the UTSA campus, and a second phase where a network of four double-gauge platforms within one radar cell is installed in the Government Canyon State Park, near San Antonio. The preliminary results indicate a correlation coefficient of 85% with a p-value of 0.0001 and a radar underestimation by 23% for paired gauge-radar events within a 6-month period (August 2006 to March 2007). Rain gauge rainfall data collected from the Park (August 2007 to June 2008) did not show a large spatial variation of rainfall distribution, which indicate most of the rainfall events in study area and time period are uniform events, although it is indicated that the "SE' of the radar cell has relative higher rain rates at certain storms and lower correlations to other three pairs of gauge measurements.

In the second part, a network of 50 rain gauges rainfall in the Upper Guadalupe River Basin were used to compare the DSP rainfalls from two radars, the KEWX radar at New Braunfels, Texas and the KDFX radar at Laughlin Air Force Base near Del Rio, Texas, for the period of September 2006 to May 2007. The rainfall data comparisons were examined based on different time scales: 6 minutes, 30 minutes, one hour, and storm total, different distances from radar to gauges (from near range, middle range to far range), and different gauge elevations from 200 m to 700 m. It is found that there is a strong radar range dependence as previously found: underestimate in the near range (< 50 km) and far range (>150 km), while overestimate in the middle range (50-150 km). The results found that at the storm total time scale that the DPS product offers a good rainfall estimation and detection, and that the correlation with rain gauges is the highest (KEWX R²= 0.59 and KDFX R²= 0.37); and the mean relative differences between gauges observations and radars estimates are the lowest (52% for the KEWX and 59% for the KDFX). It is found there is no gauge elevation dependence in the study area.

The probability of rainfall detection (POD) is dependent on time scales. It is found that radar PODs are less than gauge PODs in the 6 minutes, 30 minutes and hour time scales, while in the storm total scale, the radar PODs are higher than the gauge PODs. It is 92% for the KEWX and 85% for the KDFX at the storm scale, which are higher than the rain gauges PODs of 84% and 79%, respectively. It is supposed that radar PODs should be always larger than the gauge PODs. The reason for this inconsistencies needs further discussion.

Overall, it is clear that the DSP product is range dependent; its rainfall estimation carries some errors due to the Z-R relationship power law. This product can be used for storm monitoring and flash flood warning. A longer study period with a different rain gauges network is needed to provide further comparison and analysis between radar DSP product and rain gauges rainfall observations.

Studies on surface hydrology can generally be classified into two categories, observation for different components of surface water, and modeling their dynamic movements. This study only focuses on observation part of surface water components: snow cover, soil moisture, and precipitation. Moreover, instead of discussion on the detailed algorithm and instrument technique behind each component, this dissertation pours efforts on analysis of the standard remotely sensed products and their applications under different settings.

First in Chapter 2, validation of MODIS Terra 8-day maximum snow cover composite (MOD10A2) in the Northern Xinjiang, China, from 2000-2006, shows that the 8-day MODIS/Terra product has high agreements with in situ measurements as the in situ snow depth is larger or equal to 4 cm, while the agreement is low for the patchy snow as the in situ snow depth less than 4 cm. However, the cloud blockage in the 8-day products is still high in most winter times although the clouds are much lower than that in the daily snow cover product. According to the in situ observation, this chapter develops an empirical algorithm to separate the cloud-covered pixels into snow and no snow. This separation generates a new snow cover time series that corrects the dramatic decrease of raw snow cover area caused by MODIS cloud mask, thus yielding a better estimation of the actual snow coverage in a watershed. This solves the problem of cloud blockage frustrating hydrologic modelers. This new cloud free snow cover time series is further used to study the seasonal and inter-annual variation of snow cover at this region. Variation of snow area extent (SAE) at Northern Xinjiang is closely associated with air temperature. The increase of elevation generally accompanies the decrease of air temperature, resulting in more snow cover extent and longer snow cover duration. During the six hydrologic years from 2000-2001 to 2005-2006, the SAE has a similar pattern, although there is variation at the beginning of snow accumulation and at the end of snow melting in different years. Because of the short duration of MODIS data, the change trend of snow cover is not obvious. Therefore, the continued long-term production of MODIS-type snow cover product is critical to assess water resources of the study area, as well as other larger scale global environment monitoring, offering critical inputs for hydrologic and climatic modeling, forecasting, and climate change analyses.

Terra and Aqua satellites carry the same MODIS instrument and provide two parallel MODIS daily snow cover products at different time (local time 10:30 am and 1:30 pm, respectively). Chapter 3 develops an algorithm and automated scripts to combine the daily MODIS Terra (MOD10A1) and Aqua (MYD10A1) snow cover products, and to automatically generate multi-day Terra-Aqua snow cover image composites, with flexible starting and ending dates and a user-defined cloud cover threshold. The daily combination product (MODMYD10DC) offers much more open space than the daily MODIS Terra or Aqua only since cloud keeps moving in most times. The multi-day Terra-Aqua composite product (MODMYD10MC) provides about 3 times of images as and similar accuracy with the standard 8-day composite products with the 2003-2004 hydrologic year at test sites of Northern Xinjiang, China, and at the Colorado Plateau, USA. The new multi-day composites are significant contributions and complement to the standard MODIS snow cover product series. This approach can also be extended to other optical remote sensing images, e.g., to combine other Terra and Aqua MODIS products, like vegetation indices, etc.

Chapter 4 systematically compares the difference between MODIS Terra and Aqua snow cover products within a hydrologic year of 2003-2004, validates the MODIS Terra and Aqua snow cover products using in situ measurements in Northern Xinjiang, and compares the accuracy among the standard MODIS Terra and Aqua snow cover products, and the new combined daily and multi-day composite from both MODIS Terra and Aqua daily products. Taken the MODIS/Terra daily and 8-day products as references when being compared with MODIS/Aqua daily and 8-day products (MYD10A2), respectively, the agreement of land classification from MODIS Terra and Aqua daily and 8-day snow cover products is close to 100% in the entire year under clear sky in both morning and afternoon. In contrast, the agreement of snow classification from MODIS Terra and Aqua is high only in the winter months, decreasing in the rest of the year. The snow cover agreement in the daily products is higher than that in the 8-day products. When being compared with the in situ snow depth observations, under the so-called clear sky conditions after removing the cloud data pairs, three daily products (MOD10A1, MYD10A1, and MODMYD10DC) have similar accuracy of snow and land classification. Moreover, in the actual weather/cloud conditions, the daily combination of MODIS Terra and Aqua reduces the cloud blockage and improves the land and snow classification accuracy, although it still has high cloud percentage. Three composite products (MOD10A2, MYD10A2, and MODMYD10MC) also have similar accuracy of snow and land classification after removing the cloud data pairs. Similarly, in the actual weather/cloud conditions, the new multi-day composite product reduces the cloud blockage and improves the snow classification accuracy against either the 8-day MODIS Terra or Aqua snow cover product.

In Chapter 5, utilizing the new cloud-low multi-day composite of MODIS Terra and Aqua snow cover products, several new methods are developed to study the spatiotemporal variation of snow cover conditions from different aspects at the Northern Xinjiang and on the central Tianshan Mountains, mainly in China, partly covering Kazakhstan and Kyrgyzstan. MODIS-derived snow covered days (SCD) has good agreement of 90% with in situ SCD from hydrologic years of 2001-2005. Overall, SCD increases as elevation increases when elevation is less than 4500 m. The R² value between SCD and DEM is 0.62 on the entire Central Tianshan Mountains. Snow cover index (SCI) value contains both duration and extent information and is easily used to quantify the variation of the overall snow cover situation. MODIS-derived snow cover onset dates (SCOD) and snow cover melting dates (SCMD) also have good agreement with the major snow cover duration from in situ observation by one week forward and backward shifts because of transient snowfall events in the early and at the end of the snow season, respectively. In the six years, the perennial snow in August 2005 has the least spatial extent (2386 km²), which is believed to be a most close map of the glacier distribution at this peak area. More accurate map of glacier distribution can be obtained using Landsat or higher resolution images.

Secondly, Chapter 6 investigates the feasibility to indirectly map root-zone soil moisture using optical remote sensing techniques and in situ measurements. Specifically, covariation of root-zone soil moisture with the normalized difference of vegetation index (NDVI) from MODIS observation is studied at three sites (New Mexico, Arizona, and Texas). The three sites represent two types of vegetation (shrub and grass) and two types of climate conditions: arid/semi-arid (New Mexico and Arizona) and humid (Texas). Results show that the root-zone soil moisture has significant linear correlation with vegetation (NDVI). Then, a linear regression model is developed to estimate the root-zone soil moisture based on NDVI and in situ soil moisture. The time-series of root-zone soil moisture estimated by NDVI using the linear regression model accounts for 42-71% of the observed soil moisture variations for the 3 sites. Thus the point ground measurement of soil moisture can be interpolated to areal soil moisture using NDVI as a proxy of soil moisture variation.

Finally, Chapter 7 validates and compares the NEXRAD Stage III and MPE precipitation products using a high density rain gauge network on the Upper Guadalupe River Basin of the Texas Hill Country in 2001 and 2004. Because of the point-area representativeness error between rain gauge and radar rainfall estimation, this chapter develops a new method to automatically select uniform rainfall events based on coefficient of variation criteria of 3 by 3 radar cells. Only gauge observations of those uniform rainfall events are used as ground truth to evaluate radar rainfall estimation. Results show that rain gauge observations of uniform rainfall better represent ground truth of a 4 x 4 km² radar cell than non-uniform rainfall; MPE has better performance than Stage III; Stage III tends to overestimate precipitation (19.5%), but MPE tends to underestimate (-7.2%).

This study focused on the water-frost/ice and carbon dioxide frost interactions in impact craters of the northern seasonal polar ice cap on Mars. Studying the interaction between water and carbon dioxide frost/ice within the outer permanent ice cap can help us in determining the role of these two materials in the hydrological and geological cycles, as well as the possible biological cycle (if any) on Mars. Using THermal EMission Imaging System (THEMIS) infrared and visible images, Thermal Emission Spectroscopy (TES) Lambert albedo and target temperature data and Mars Orbital Laser Altimeter (MOLA), a detailed study of seven previously identified craters with high albedo events (HAEs) in the period of late spring to late summer was carried out. Results indicate six of the seven observed impact craters with both AM and PM HAEs are associated with a permanent water-ice body (or patch) covering a part or all of those crater floors, while one impact crater with AM-only HAEs is not associated with a permanent water-ice body. This observation is consistent with a conceptual model proposed by Xie et al. [in press] that an exposed permanent water-ice body is the basis for the AM and PM HAEs in those craters, while a subsurface water ice body or ice-rich regolith in the near subsurface is a necessary condition for the AM-only HAEs crater. Variations in albedo in Phases 2 and 3 of the Martian annual cycle are found in some craters. This is mainly due to the exposed, permanent water-ice body (patch) that is only a small portion of the crater floor, resulting in a limited contribution from the water-ice body to the albedo change of the entire crater floor measured from the TES instrument.

Satellite remote sensing provides spatially and temporally continuous datasets for vegetation analysis at large coverage areas. In this study, three vegetation indices and band composites derived from Moderate Resolution Imaging Spectroradiometer (MODIS) reflectivity product (MOD9A1: 8-day and 500m) are examined for their correlations with the Remote Automated Weather Stations (RAWS)-derived live fuel moisture content (LFMC) during the time period of February 2000 to December 2004. The correlation analyses were based on green-up months (April 1 – October 31) between collocated LFMC from a RAWS station and MODIS-derived vegetation indices or band composites from one pixel where the RAWS station is located at the center (near center) of the pixel. Three RAWS together with additional 4 sites (total 7 pixels) were chosen in the West Texas study area. In situ data were also analyzed to determine the best fit for correlation with MODIS vegetation indices. It is found that, based on RAWS-derived LFMC, the multiple variable regression (7 bands together) approach results in the best correlation in establishing equations to produce LFMC maps for all study sites for the mixed vegetation community of West Texas. The strongest relationship between RAWS and MODIS was found at Fort Davis site3 in multiple regression analysis (r² = 0.55, p < 0.0001). The best correlation (r² = 0.61, p < 0.0001) between MODIS indices and in situ measurements was based on Normalized Difference Infrared Index band 6 (NDII₆) at the Fort Davis Mountains within the High Pocosin vegetation fuel class.