WIND EROSION EVALUATION BY DIGITAL IMAGE PROCESSING 
Stella Maris Navone 
Universidad Buenos Aires 
Facultad de Agronomia 
Departamiento de Suelos - Manejo e Conservaciön 
Congresso 1566 
Buenos Aires, 1428, Argentina 
ISPRS Commission VII / Working Group 6 
ABSTRACT 
Increasing desertification has been observed for a long time in the Abaucán River basin (7535 km2) with the 
Fiambalá and the Tinogasta valleys in Catamarca province, and the Miraflores River basin (2853 km2) in 
Jujuy Puna (Argentine Republic). Our study in the field indicate that several factors and processes contribute 
to land degradation in both areas although from the point of view of their quantitative importance, there is a 
main process: wind accelerated erosion. To understand what happens, it's very important to have an 
appropriate tool to make a proper diagnosis. The objective of this research was to evaluate the effectiveness 
of the Landsat satellite image in differentiating among degree of wind erosion. Two sets of MSS and TM 
data were used in each of the three areas, a dry season data (winter) and rainny season data (summer). 
Jeffries-Matusita distance (Swain and Davies, 1978) was applied on to choose the best band combination. To 
enhance wind erosion features, digital processes were tried. Erosion was measured in the field according to 
the "Desertification Method" (FAO, 1984). The results showed that soil loss caused by wind erosion was 
detectable in 1. very severe degree, when histogram equalization was applied 2. severe degree in the image 
enhanced with Principal Components 3. moderate degree low filtering transformation of the data and with 
nonlinear enhancements. Dry season (winter) is the best moment to carry out wind erosion surveys by MSS 
and TM satellite images. 
090