Integrated modeling and remote sensing of soil moisture
暂无分享,去创建一个
P. C. D, MILLY & Z. J. KABALA Department of Civil Engineering, Princeton University, Princeton, New Jersey 08544, USA Abstract Modeling and remote sensing are two complementary procedures that yield information on the moisture content of the soil beneath the land surface. Each has inherent errors. It is reasonable to expect that a combination of modeling and remote sensing would yield more accurate estimates of soil moisture content and related variables than can be obtained by either approach alone. We propose that the problem of integrating models and remote sensing be approached using the extended Kalman filter (EKF). The EKF integrates uncertain state dynamics and uncertain measurement models to arrive at an optimal estimate of the system states. The EKF also yields an explicit measure of the accuracy of the state estimates. In a specific application, we demonstrate the use of the EKF to determine moisture content beneath a bare soil. The use of EKF provides estimates of soil moisture superior to those obtained separately by either remote sensing or modeling alone. Background The amount of water in the top meter of the earth's soil is a critical variable that controls a number of hydrologie, biological and meteorological processes (Schmugge et al., 1980). Soil moisture in the top few centimeters governs the partitioning of rainfall into infiltration and surface runoff, the latter being the major source of flood flow in many streams and rivers. The rate of evaporation of water from soil is strongly dependent upon the amount of moisture near the soil surface. Transpiration and growth of crops and other plants depend upon the availability of water to the plant roots. Because a significant quantity of heat is absorbed when vaporization of liquid water occurs, the rates of evaporation and transpiration have a significant influence upon the energy balance of the surface of the earth. Therefore, soil moisture affects weather and climate by controlling the fluxes of water and sensible heat into the atmosphere. Finally, we note that the water yield of a catchment decreases in direct relation to evaporation and transpiration. The analysis of many of the above-mentioned processes is often complicated by the variability in time and space of the moisture content of the soil. Variations in time result from the dynamic nature of the hydro-logical cycle, with periodic forcing at the annual and daily scales and with more random forcing due to the passage of weather systems. Variations in depth are intimately related to the temporal fluctuations in forcing. Generally, surface conditions respond to all frequencies of forcing. Deeper soil moisture responds more slowly, due to smoothing by the intervening surface layers. In order to deal with the problems of soil moisture variability, analyses of the many processes dependent upon soil moisture have often 331
[1] Arthur Gelb,et al. Applied Optimal Estimation , 1974 .
[2] T. Schmugge,et al. Survey of methods for soil moisture determination , 1980 .
[3] P. Milly. Moisture and heat transport in hysteretic, inhomogeneous porous media: A matric head‐based formulation and a numerical model , 1982 .
[4] P. Milly,et al. A Simulation Analysis of Thermal Effects on Evaporation From Soil , 1984 .
[5] W. G. Gray,et al. Finite Element Simulation in Surface and Subsurface Hydrology , 1977 .