2008 Joint Meeting of The Geological Society of America, Soil Science Society of America, American Society of Agronomy, Crop Science Society of America, Gulf Coast Association of Geological Societies with the Gulf Coast Section of SEPM

Paper No. 11
Presentation Time: 10:30 AM

Mapping of Spatio-Temporally Variable Semi-Arid Soils Using Optical Imagery


HENDRICKX, Jan M.H.1, ENGLE, Emily1, HARRISON, J. Bruce J.1, BORCHERS, Brian2, HONG, Sung-ho1 and FLEMING, Kathy1, (1)Earth & Environmental Science, New Mexico Tech, Socorro, NM 87801, (2)Mathematics, New Mexico Tech, Socorro, 87801, hendrick@nmt.edu

Mapping of spatio-temporally variable semi-arid soils is of critical importance for military geology since soil conditions affect virtually all aspects of Army activities. Soil conditions affect operational mobility, detection of improvised explosive devices, natural material penetration, military engineering activities, blowing dust and sand, watershed responses, and flooding. Unfortunately, soil maps in semi-arid regions either do not exist at all or have been made at scales that are too large for many Army applications. The main reason is that the economic value of semi-arid soils is so low that investments in soil mapping are not justified with the exception of some high-value agricultural soils. The objective of this contribution is to present recent progress with our approach to digital mapping of semi-arid soils using optical images of Landsat for analysis with SEBAL (Surface Energy Balance Algorithms for Land) and METRIC (Mapping Evapotranspiration with Internalized Calibration). For each pixel in an image these algorithms allow the quantification of ecohydrological parameters such as Normalized Difference Vegetation Index, albedo, surface temperature, evapotranspiration and root zone soil moisture for high-resolution classification of soil types. Contrary to conventional perception that optical imagery only provides information about the surface of each pixel, using SEBAL/METRIC we can investigate the soil conditions of the subsurface occupied by roots, which is an insight that other studies have not used. Permanent spatial distributions of soil characteristics are mapped by pooling all information available in ten or more images that characterize soil differences during a wide range of soil moisture conditions. Temporal variability of soil characteristics is analyzed by making comparisons between single images obtained at different dates. Our new approach is tested in semi-arid desert and riparian areas located in the Middle Rio Grande Valley of New Mexico.