Clasificación y predicción de cambio de cobertura de suelo de la cuenca del Río Paute utilizando herramientas geoinformáticas

Abstract

This study aims to analyze, classify and predict the changes that have taken place in land cover Advertise River Basin (CRP) for the realization of this work has been proposed in order to classify the land cover using geoinformatics techniques taking as starting image 1987 which was graded and designed in conjunction with the image, 2001, later to be verified with real information from the classified image 2010. the images involved in this research were acquired from the Landsat 5 (sensor TM) and Landsat sensor 7 (ETM + sensor), these images were classified using a hybrid method of classification based on pixels and segments. Changes, exchanges and systematic transitions involving different land cover were calculated: Forest / vegetation, pasture / crop areas, population centers, bodies of water, Paramo Ground bare / sparsely vegetated areas; stochastic methods using for this prediction. GIS software IDRISI Selva was used for prediction, classification methods using rigid, multi-criteria analysis and cellular automata integrated with Markov chains. The validation of the classifications and model prediction of confusion matrices employees with levels of agreement obtained by above 80% reliability and concordance kappa. At the end of this research it was demonstrated increases in 5.07% of soil degradation through 2010 and an estimated increase of 6.32% over a period of 15 years, reduction of principal land cover (forests / vegetation), the displacement of the agricultural frontier into a 2.94% because of the urban sprawl that has increased over the past 23 years 0.96%