Land Use and Land Cover Change Modeling Using Remote Sensing and Soft Computing Approach to Assess Sugarcane Expansion Impacts in Tropical Agriculture

Abstract:

Agriculture is related with land-use/cover changes (LUCC) over large areas and, in recent years, increase in demand of ethanol fuel has been influence in expansion of areas occupied with corn and sugar cane, raw material for ethanol production. Nevertheless, there´s a concern regarding the impacts on food security, such as, decrease in areas planted with food crops.

Considering that the LUCC is highly dynamic, the use of Remote Sensing is a tool for monitoring changes quickly and precisely in order to provide information for agricultural planning.

In this work, Remote Sensing techniques were used to monitor the LUCC occurred in municipalities of São Paulo state- Brazil related with sugarcane crops expansion in order to (i) evaluate and quantify the previous land cover in areas of sugarcane crop expansion, and (ii) provide information to elaborate a future land cover scenario based on Self Organizing Map (SOM) approach.

The land cover classification procedure was based on Landsat 5 TM images, obtained from the Global Land Survey. The Landsat images were then segmented into homogeneous objects, with represent areas on the ground with similar spatial and spectral characteristics. These objects are related to the distinct land cover types that occur in each municipality. The segmentation procedure resulted in polygons over the three time periods along twenty years (1990-2010). The land cover for each object was visually identified, based on its shape, texture and spectral characteristics. Land cover types considered were: sugarcane plantations, pasture lands, natural cover, forest plantation, permanent crop, short cycle crop, water bodies and urban areas.

SOM technique was used to estimate the values for the future land cover scenarios for the selected municipalities, using the information of land change provided by the remote sensing and data from official sources.