|
Description:
|
The dissertation focuses on characterization of subpixel variability within a
satellite -based remotely sensed coarse -scale soil moisture footprint . The underlying
heterogeneity of coarse -scale soil moisture footprint is masked by the area -integrated
properties within the sensor footprint . Therefore , the soil moisture values derived from
these measurements are an area average . The variability in soil moisture within the
footprint is introduced by inherent spatial variability present in rainfall , and geophysical
parameters (vegetation , topography , and soil ) . The geophysical parameters /variables
typically interact in a complex fashion to make soil moisture evolution and dependent
processes highly variable , and also , introduce nonlinearity across spatio -temporal scales .
To study the variability and scaling characteristics of soil moisture , a quasi -distributed
Soil -Vegetation -Atmosphere -Transfer (SVAT ) modeling framework is developed to
simulate the hydrological dynamics , i .e . , the fluxes and the state variables within the
satellite -based soil moisture footprint . The modeling framework is successfully tested
and implemented in different hydroclimatic regions during the research . New multiscale data assimilation and Markov Chain Monte Carlo (MCMC ) techniques in conjunction
with the SVAT modeling framework are developed to quantify subpixel variability and
assess multiscale soil moisture fields within the coarse -scale satellite footprint .
Reasonable results demonstrate the potential to use these techniques to validate
multiscale soil moisture data from future satellite mission e .g . , Soil Moisture Active
Passive (SMAP ) mission of NASA . The results also highlight the physical controls of
geophysical parameters on the soil moisture fields for various hydroclimatic regions .
New algorithm that uses SVAT modeling framework is also proposed and its
application demonstrated , to derive the stochastic soil hydraulic properties (i .e . , saturated
hydraulic conductivity ) and surface features (i .e . , surface roughness and volume
scattering ) related to radar remote sensing of soil moisture . |