Soil Degradation Evaluated by a 27 years Landsat image (Vis-Nir-Swir-Tir), climate and digital elevation derivatives
Abstract
According to Food and Agriculture Organization of the United Nations, 30% of the global soils are degraded. Therefore, novel researches on soil degradation process are imperative to prevent damages on social and environmental dynamics. Since we have a wide world dimension, and few manpower, we have to focus on high dimensional evaluation techniques such as remote sensing. The main goal of this work was to develop a method, based on a 27 years time-series of satellite images (Landsat), from which determine the most important factors on soil degradation. The area is located in south Brazil with a 1400 km2 area. The steps of the method are as follows: a) we collected images from the area and based on a novel technique determined the areas with exposed soils; b) we quantified soil properties such as clay and capacity of ionic exchange based on pixel spectra signature; c) the technique also indicated how many times a single pixel was with bare soil during the period; d) we also determined the surface temperature based on band 6; e) using elevation model we created the layers LS factor, drainage density, topographic wetness index, solar radiation; f) we also determined climate information (water balance); g) organic matter (OM) was also estimated. All factors from item a to f were balanced and overlapped (GIS) to generate an index of soil degradation, SD (fig 1a) - values from 1 (low risk) to 5 (high risk). We concluded that 30% of the area is degraded. SD presented coherent values with OM and validate the method. We observed that areas with higher SD (5) contain 43.6% less OM than the ones with low risk (1). In addition, the soil spectral reflectance curve was analyzed concluding that degraded soils shows higher intensity. The current land use (fig 1b) was correlated demonstrating that a higher risk of SD happens mainly in sugar cane (41.6%) in contrast to pasture (16.9%) and forestry (11.7%). Therefore, this approach allows land uses decision-making and public policies.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2017
- Bibcode:
- 2017AGUFM.B31D2016D
- Keywords:
-
- 0439 Ecosystems;
- structure and dynamics;
- BIOGEOSCIENCES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1622 Earth system modeling;
- GLOBAL CHANGE;
- 1640 Remote sensing;
- GLOBAL CHANGE