H33E-0508 1340h
EFFECT OF SCALING TRANSFER BETWEEN EVAPOTRANPIRATION MAPS DERIVED FROM LANDSAT7 AND MODIS IMAGES
Remotely sensed images of the Earth's surface provide information about the spatial distribution of evapotranspiration. Since, the spatial resolution of evapotranspiration predictions depends on the sensor type; scaling transfer between images of different scales needs to be investigated. In this study, the Surface Energy Balance Algorithm for Land (SEBAL) was used to derive evapotranspiration maps from Landsat7 Thematic Mapper (TM) and Moderate Resolution Imaging Spectroradiometer (MODIS) images. LandSat7 has a spatial resolution of 60 m, MODIS of 1000 m. Two up-scaling procedures are evaluated. The first consists of averaging 60 by 60 m LandSat pixels of spectral radiances or of intermediate SEBAL input surface parameters (albedo, NDVI, surface temperature) to obtain 1000 by 1000 m pixels at the MODIS scale before SEBAL is applied. The second consists of first applying SEBAL and then to average from 60 m to 1000 m spatial resolution. The averaging process (aggregation) will include calculating arithmetic and geometric means. In the downscaling process (MODIS to LandSat), an earlier LandSat image will be used to characterize the fine scale variability within the large MODIS pixels. Two down-scaling procedures are evaluated. The first consists of downscaling the original MODIS spectral data; the second of downscaling the evapotranspiration maps at MODIS scale. The objective of this study is to investigate the effect of the scaling transfer processes on the spatial distribution of evapotranspiration in arid riparian areas.
H33E-0509 1340h
Lidar Measurement of Boundary Layer Evolution to Determine Sensible Heat Fluxes
The estimation of large scale fluxes of heat and water vapor is crucial for a number of reasons, including weather prediction, climate, agriculture, and water resources management. It is known that these fluxes are not uniform and that the distribution of the surface heat fluxes is a major factor in producing and modifying mesoscale atmospheric flows, turbulence and evaporation. Local variability in the values is the result of variations in soil type, moisture content, type of vegetative cover, and wind speed, among other factors. Obtaining high quality data requires expensive equipment that needs periodic maintenance and attention. So while point sensors cannot be extensively deployed because of cost, conventional measurements represent the surface fluxes over very limited areas. This work evaluates a method where, with a relatively simple, vertically staring lidar, larger scale estimates of the sensible heat flux can be inferred. The estimation of a spatially integrated sensible surface heat flux and the determination of regional heat fluxes over heterogeneous land surfaces is needed to address a number of problems concerning long-range and mesoscale transport models of pollutants and parameterizations in climate models.
H33E-0510 1340h
Root Zone Soil Moisture Assessment Using Passive Microwave Remote Sensing and Distributed Hydrologic Modeling
Soil moisture is a fundamental state variable and it varies spatially due to topography, soil, precipitation variability and landuse-landcover, and temporally, due to difference in hydrologic characteristics and controls. Estimation of profile soil moisture using remotely sensed land surface moisture data with the combination of forward soil hydrologic modeling is very promising. This integrated method may become resourceful solution to the problem for profile soil moisture estimation and its transient behavior. In a preliminary attempt to understand soil moisture land surface dynamics, an effort is being made to assess soil moisture on the watershed scale. A method has been implemented to combine hydrologic model and passive microwave land surface soil moisture observation to predict root zone soil moisture. The prime focus of this work is to combine HYDRUS-1D model with Soil Survey Geographic (SSURGO) Database (30 x 30 meters), NEXRAD based precipitation, LANDSAT7 landcover classification and ESTAR derived surface soil moisture in a GIS platform. We applied and tested this integrated approach in the Little Washita watershed during SGP97 using ground and remotely sensed data set for a month period. The integrated model facilitates to identify the critical parameters that control the spatio-temporal variability of the soil moisture fields and a good assessment of soil moisture in the root zone.
H33E-0511 1340h
High-Resolution Three-Dimensional Forward Simulations of Flow Into Heterogeneous Unsaturated Media: Comparison of Simulated and Geophysical Data
High-resolution subsurface flow and transport simulations using PorSalsa, a massively-parallel, two-phase, three-dimensional finite element flow code, are performed and compared to data from the Sandia-Tech Vadose Zone Facility (STVZ). The STVZ is a state-of-the-art infiltration facility where a variety of hydrologic sensors and geophysical imaging techniques are employed for monitoring water flow and salt tracer transport through highly heterogeneous fluvial sediments. Current research at the STVZ involves an assessment of electrical resistivity tomography (ERT) and cross-well ground penetrating radar (XGPR) measurement resolution as employed at the STVZ. We present and discuss results from high-resolution unsaturated flow simulations based on the STVZ infiltration test that play a vital role in this research. These simulations employ layered homogeneous and heterogeneous hydraulic property fields, where the hysteretic hydraulic properties have been measured on samples from the STVZ using standard laboratory techniques, and GSLIB is used to generate geostatistically correlated heterogeneous fields of porosity, permeability, and air entry values. We will discuss the effects of grid resolution, up-scaling techniques for the heterogeneous fields, and the impact of uncertainty in the measured properties on the simulations. We will also present and discuss the differences between the simulated results and measured field data from hydrologic/geophysical sensors.
H33E-0512 1340h
Forward and Inverse Simulations of Electrical Resistivity Tomography to Demonstrate the Resolution of Fine-Scale Hydrological Features Within Heterogeneous Unsaturated Media
Converting hydrological parameters from very high-resolution flow simulations to electrical conductivity created synthetic electrical resistivity tomography data. The data were then forward modeled using very fine, high-resolution meshes and inverted using the same coarse meshes that were used for normal inversion of field data. The goal was to simulate earlier results from the Sandia-Tech Vadose Zone Facility, a state-of-the-art infiltration facility where geophysical imaging techniques were employed to monitor the unsaturated flow of potable water and transport of a salt tracer through fluvial sediments. We discuss the relative resolution of the ERT compared to the original hydrological models. We also discuss the effects on resolution of mesh design and the values assigned to various modeling parameters. Finally we compare the optimized parameters of this study with those used during the field program.
H33E-0513 1340h
Combining Cross-Borehole Georadar and Electrical Resistance Tomography to Monitor Moisture Content and Pore Water Electrical Conductivity in the Vadose Zone
Cross-borehole georadar and electrical resistance tomography (ERT) measurements are being conducted at two field sites in Denmark to monitor spatial and temporal variation of moisture content and tracer transport. At both field sites the geology is characterised by a thick layer (i.e.~20-30 m) of unsaturated melt water sand and gravel. At one of the field sites, Hjelm Hede, a very simple setup was constructed. Four boreholes were drilled along a line to a depth of 12 m. The outer two boreholes (7 m apart) were equipped with ERT instrumented PVC-tubes (electrodes every 50 cm) while the inner boreholes (5 m apart) had access tubes for georadar installed. The estimated water content values collected by the two large-scale geophysical methods have at this locality been compared to water content values determined by a conventional method - TDR. During previous investigations a total of 7 one-meter-long TDR probes were installed horizontally at five depths, from 0.75 m to 6.00 m. The TDR probes were installed in a well located approximately 30 m from the test area. The use of TDR probes to determine water content represents the soil physical properties at small-scale. A comparison will therefore be appropriate when evaluating properties across different sampling scales. Initial field results based on georadar match TDR results within a few percent. The second field site, Arrenaes, is located by an artificial infiltration plant. Two identical field setups have been established each having four ERT and four georadar boreholes. The boreholes were drilled to form a cross consisting of two lines, each with the same geometry as was used at Hjelm Hede. The two setups are located 8.5 m apart enabling an evaluation of the spatial variability. In future tests one of the fields will, furthermore, be irrigated to accelerate flow. The setups provide the possibility of combining the water content images using cross-borehole georadar with the bulk conductivity images achieved from the cross-borehole ERT to monitor the variations in fluid conductivity. These changes may arise as a result of natural changes, i.e.~changes in temperature and/or rainfall conductivity, or artificial changes, such as an applied tracer.
H33E-0514 1340h
Spatial Variability of Electrical Conductivity and Its Relationship to Soil Properties at The University of Mississippi Soil Moisture Observatory
The Soil Moisture Observatory (SMO) at the University of Mississippi (UM) is a 5 acre tract of a former agricultural field at the UM Biological Field Station. Preliminary investigations of this site included 60 continuous soil cores using the Geoprobe sampling technique. These soil cores were taken to a depth of 1.5 meters to correspond to the approximate depth of penetration for a Geonics EM38. The Geonics EM38 uses electromagnetic induction to measure apparent electrical conductivity (ECa) of the soil. Weekly readings using the EM38 were conducted at the 60 sites. Soil samples were collected on the fourth week of EM38 readings. These samples were analyzed for particle size distribution, porosity, bulk density, iron content, and volumetric moisture content. This analysis determined that soil physical and chemical properties control the spatial variations in ECa. Volumetric moisture content shows correlation lengths less than 25m at shallow depths and greater than 35m deeper; variability also increases with depth. Volumetric moisture content appears to be most closely related to soil particle size. The temporal variations of the variograms of ECa indicate a complex relationship between soil properties and ECa.
H33E-0515 1340h
Comparing Different Estimates of Evaporation Coefficient From Two-Stage Evaporation of Soil Surfaces
Evaporation coefficient ({\it K}$_{E}$) allows for estimation of soil evaporation without in situ measurements. In this study, a resistance based method ({\it K}$_{E}$$^{r}$) and a temperature based method ({\it K}$_{E}$$^{t}$) for estimating {\it K}$_{E}$ were compared with high-frequency measurements of two-stage soil evaporation. An alternative method ({\it K}$_{E}$$^{a}$) for estimating {\it K}$_{E}$ was presented in which soil surface resistance ({\it r}$_{s}$) was computed with temperature differences between surface and overlying air ({\it T}$_{r}$-{\it T}$_{a}$). Results indicated that more accurate estimates of {\it K}$_{E}$ were found in the first stage than in the second stage. Compared to the measurements, the {\it K}$_{E}$$^{r}$ method produced the best estimates with a correlation coefficient ({\it r}$^{2}$) of 0.88, a root-mean-square-difference (RMSD) of 0.04, and a mean-absolute-difference (MAD) of 0.03. The {\it K}$_{E}$$^{t}$ method overestimated in both stages of evaporation with the lowest {\it r}$^{2}$ (0.30) and largest RMSD (0.14) and MAD (0.13). The {\it K}$_{E}$$^{a}$ method produced estimates of {\it K}$_{E}$ comparable to the {\it K}$_{E}$$^{r}$ method, which was a significant improvement from the {\it K}$_{E}$$^{t}$ method. The correlation coefficient between {\it K}$_{E}$$^{a}$ and measurements was 0.64, and the RMSD and MAD was 0.05 and 0.04, respectively. Future investigations should focus on improving the functional relationship between {\it r}$_{s}$ and ({\it T}$_{r}$-{\it T}$_{a}$).