HR: 09:45h
AN: H11A-08    [PDF]
TI: Spatial Variability and Multiscale Errors in Assimilated Soil Moisture Fields
AU: * Chintalapati, S
EM: chintala@uiuc.edu
AF: Dept. of Civil and Environmental Engg. University of Illinois, 205 N. Mathews Ave., Urbana, IL 61801 United States
AU: Kumar, P
EM: kumar1@uiuc.edu
AF: Dept. of Civil and Environmental Engg. University of Illinois, 205 N. Mathews Ave., Urbana, IL 61801 United States
AB: Coupled land-atmosphere models are increasingly focused on using assimilated near-surface soil-moisture to improve the prediction of moisture and heat fluxes. Given that these models have a non-linear dependence on the soil-moisture state, it should be expected that errors in the observed soil-moisture will propagate through the system and influence the predictions of the model. Of particular concern is the situation when model predictions are made at scales that are different from that of the observation scales of the soil-moisture. In this case the transformation process of the soil-moisture from one scale to another introduces additional errors that will influence the prediction. The purpose of this study is to develop an understanding of the influence of multiscale observational errors in the soil-moisture on the prediction of surface fluxes. The following approach is adopted. We use the SGP'97 near-surface soil-moisture observation from ESTAR images at 0.8 km (resampled to 1 km). These observations are available for 16 days during the period June 18, 1997 to July 16, 1997. Using these we obtain estimates of the near-surface soil-moisture at several scales (1, 2, 4, 8, 16 and 32 km). This is accomplished through a multiscale estimation technique using a mean differenced fractal model [Kumar, 1999]. These multiscale near-surface observations are assimilated using extended Kalman filtering algorithm embedded in the NCAR-LSM (Land Surface Model). The assimilation is performed, once daily on the days when observations are available, at all scales for the entire study period to predict the soil-moisture profile and the surface energy fluxes. This allows us to address two key issues: (1) assessment of how the estimation error evolves with different moisture conditions as a function of scale, and (2) assessment of the spatial variability of assimilated fields at different scales. From the multiscaling results, we can observe that the estimation errors grow as the moisture increases in general at all the scales. Also there is a decrease in the estimation error as the scale increases. The assimilated fields show consistent spatial variation at all scales but also show their dependence on assimilation frequency and rainfall events. This analysis will help us understand the model response as a function of scale. It also helps in developing a better framework for specifying the error statistics in the assimilation algorithm. [Kumar, 1999] Kumar, P., A Multiple Scale State-Space Model for Characterizing Subgrid Scale Variability of Near-Surface Soil Moisture, IEEE Transactions on Geoscience and Remote Sensing, 37(1), January 1999.
DE: 1640 Remote sensing
DE: 1655 Water cycles (1836)
DE: 1860 Runoff and streamflow
DE: 1866 Soil moisture
SC: Hydrology [H]
MN: 2003 Fall Meeting