Characterization of the spatial-temporal variability of soil moisture by remote sensing /
Characterization of spatial and temporal variabilities of soil moisture, spectral formalism of soil moisture estimation and sampling error simulation study were conducted to understand soil moisture field and to establish global monitoring strategy. Linear relation between soil moisture and porosity...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1999.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://search.proquest.com/docview/304563313?accountid=7082 |
| Summary: | Characterization of spatial and temporal variabilities of soil moisture, spectral formalism of soil moisture estimation and sampling error simulation study were conducted to understand soil moisture field and to establish global monitoring strategy. Linear relation between soil moisture and porosity is dramatically improved with increasing pixel size although linear relation between soil moisture and soil properties is very weak. The relation between field variance and aggregation area follows power law between log scale 4 and 7. Scaling analysis indicates that the power law exponent becomes smaller with increasing area, which allows the assumption that the soil moisture field is stationary in large area. Variogram analysis shows that the stationarity of soil moisture field is changed by meteorological condition. Spectrum of soil moisture field shows there is no dominant spatial frequency. Two-dimensional correlogram of the soil moisture and brightness temperature Gelds shows strong anisotropy. Correlation structure of the soil moisture field is changed by drying or rainfall process. Average correlation length of the soil moisture consists of Long'-14km and Lat'-36km. Autoregressive exogenous model (ARX) with lag-1 correlation coefficient suggested for temporal soil moisture model. The Monsoon '90 soil moisture data indicate that diurnal cycle causes 1-4% sampling error. (5 a.m. - 9a.m. : 1%, 1 p.m. - 3 p.m. : 4% ). North-Nakamoto formalism(1989) was used to compute the sampling error for the soil moisture field estimation. The space-time discrete design filter was evaluated and it is applicable to all kinds of sampling design. Missing temporal measurements in SGP '97 soil moisture field make it difficult to estimate the spectra directly from observed record. The soil moisture spectrum was estimated using rainfall and soil moisture models tuned parameter to SGP '97 data. Estimated sampling error of daily electronically scanned thinned array radiometer (ESTAR) soil moisture sampling 2.6%. Soil moisture data were generated by using rainfall and soil moisture model that considered the diurnal cycle and topographic effects with parameters tuned to SGP '97 data. Partial coverage caused greater sampling errors than temporal gaps. The larger the capacity for holding water, the less sampling error is generated. Block random installation gets more accurate data than random installation of soil moisture gages. The greater the loss coefficient, the greater the sampling errors. Sampling error linearly increased with increasing diurnal cycle amplitude. |
|---|---|
| Item Description: | Vita. "Major Subject: Civil Engineering". |
| Physical Description: | xiv, 155 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilm Inc. |
| Bibliography: | Includes bibliographical references (leaves 137-142). |