Geophysical    Engineering
Geophysical Engineering
Future Students  |  Current Students  |  Parents  |  Alumni   |    Giving   |   Faculty & Staff   |   Visitors  |   Athletics
Xiaobing Zhou, Associate Professor of Geophysicis

Derivation  of  Liquid  Water  Content  for  Mixed  Terrain  of  Snow  and  Soil  from  Radar  Backscattering  Coefficient



ABSTRACT:

The capability of measuring the water availability in soil for crop and other vegetation from space on a large scale but also at acceptable resolution independent of the weather conditions can provide a great opportunity for precise agriculturing. Radar backscattering data has the potential of such a capability. To realize such a potential, an accurate algorithm to derive the soil moisture content and liquid water content (LWC) in snow is imperative. Liquid water in snow not only affects the snow albedo, which is an important energy balance parameter in the interaction between atmosphere and land surface, but also a good indicator of snow melting. Mixed land surface of snow and soil is a common phenomenon in the spring time in mid- to high-latitude regions. The mixed terrain occurs at the same time frame as growing season for crops and other vegetation. Knowledge of the LWC in the mixed terrain is thus useful information for crops monitoring and irrigation management. Based on Fung's Integral Equation Model (IEM) (Fung et al., 1992; Fung, 1994), Rayleigh scattering approximation, empirical soil moisture - dielectric constant model, and snow spherical model, an integrated algorithm for the retrieval of LWC in snow and soil is developed and assessed using RADARSAT ScanSAR data taking Seward Peninsula, Alaska as the test site. Integration of a module that derives the salinity in soil from radar backscattering coefficient data is being pursued.

DATA SOURCES AND METHODS:
  • Integration algorithm for LWC derivation is based on the Fung's IEM model, Rayleigh scattering approximation, empirical soil moisture - dielectric constant model, and snow spherical model;
  • Separation of snow cover from snow-free land surface is done by using MODIS 8-day product;
  • RADAR data is RADARSAT ScanSAR provided by Alaska SAR Facility;
  • DEM data is USGS GTOPO30 data.
RESULTS:
  1. (Figure 1) The integrated algorithm is shown as follows.


  2. GEOP 302

    Figure 1. Flow chart for the retrieval of LWC in snow and soil for mixed terrain.

  3. (Figure 2) The land surface of Seward Peninsula, Alaska is a mixed terrain of snow and soil from spring to early fall as observed by the MODIS.


  4. GEOP 302

    Figure 2. As observed by MODIS snow 8-day product in 2001, the terrain of Seward Peninsula, Alaska is a mixed terrain of snow and snow-free land from the end of May till the end of August, when another cycle of snow accumulation begins.

  5. (Figure 3) Liquid Water Content (LWC) is derived from the integrated algorithm (bottom panel) as snow melt proceeds (top panel). It is observed that in LWC in snow is lower than in soil. LWC in the western region is higher than the eastern region.


  6. GEOP 302

    Figure 3. LWC derived from the integrated algorithm shows (bottom panel) that LWC increases as snow melt proceeds (top panel). Western region has higher LWC for either snow cover or snow-free land surface.

  7. (Figure 4) Comparison of the LWC and snow maps with the DEM and land cover classification map from Landsat data shows that LWC in the low north-western coastal region is very high and LWC in the eastern inland region is relatively low. However, LWC in the highly-elevated mountainous regions is over-estimated, probably due to the correlation length is derived from the images only for the flat regions. Work on derivation of correlation length from mountainous region is going on.


  8. GEOP 302

    Figure 4. LWC for August 25, 2001 derived from the algorithm and snow cover from MODIS 8-day data versus DEM and land cover classification map from Landsat data.