作者: C.A. Portillo-Quintero , A.M. Sanchez , C.A. Valbuena , Y.Y. Gonzalez , J.T. Larreal
DOI: 10.1016/J.APGEOG.2012.06.015
关键词:
摘要: This work synthesizes results from the application of land cover classification techniques and probability sampling satellite imagery for estimating forest extent deforestation in Lake Maracaibo Basin (Venezuela Colombia). A map was produced using a semi-automated supervised routine on MODIS 8-day 500-m acquired January 2010. Results show that forests occupy 29,710 km2 which represents 38% basin's total terrestrial landmass. From this extent, 61% belongs to Venezuela 39% falls within Colombian region. Findings indicate drastic decrease as result anthropogenic agricultural urban expansion, especially when compared its potential ‘Maracaibo dry forests’ ‘Venezuelan Andean montane ecoregions. Using time series Landsat imagery, rates 1985–2010 period were calculated. The analysis performed 24 samples blocks 10 × randomly allocated previously defined change strata. general spatial distribution predicted by simple regression model between sample prior probabilities at basin scale. Our low (<0.5%/y) 85% basin, with highly focalized fronts (intermediate-to-high rates, <2.5%/y) three regions: a) Motatan river sub-basin Eastern Cordillera, b) lower slopes Catatumbo c) submontane regions Apon Santa Ana sub-basins. paper lead way understanding current patterns socioeconomic drivers clearing Basin. study also demonstrates feasibility alternatives methods time-consuming financially unsustainable traditionally used national sub-national scale other Latin American countries.