发帖
雷达卡
英文文献：Time Series Analysis Of Satellite Data: Deforestation In Southern Mexico-卫星数据的时间序列分析:墨西哥南部的森林砍伐
英文文献作者：Geoghegan, Jacqueline,Hewitt, Julie A.,Vance, Colin
英文文献摘要：
Tropical deforestation is significant to a range of themes that have relevance for the study of environmental change and economic development, including global warming, land degradation, species extinction, and sustainability issues. Recognition that both the location and pattern of forest clearance are often as important as its magnitude has motivated an increasing number of econometric studies that link satellite data and government census data with the aim of modeling the spatial dimensions of deforestation processes. Initial research focused on time series analysis, while recent work has started developing models that make use of time series data on land use. In this paper, we use satellite data from three dates over an approximate 15-year period to estimate the probability of a satellite pixel being in a forested or human-disturbed state. Our study focuses on land-use change in an agricultural frontier spanning the southern Mexican states of Campeche and Quintana Roo. This region contains one of the largest and oldest expanses of tropical forests in the Americas outside of Amazonia and has been identified as a "hot spot" of forest and biotic diversity loss. Over the past 30 years, these forests have been under sustained pressure following the construction of a highway in 1972 that opened the frontier to settlement. The road was part of a larger development effort to promote agricultural colonization and has contributed to a prolonged period of land transformation that has been captured by Thematic Mapper (TM) satellite imagery. We capture these landscape dynamics by assembling a spatial database that links the pixels from three TM images spanning the years 1986-1997 and other spatial environmental and GIS-location derived data with government census socio-economic data of data. We develop a simple utility-maximizing model of the forest clearance decision. Based on previous research, the theoretical model suggests many possible determinants of forest clearance in an economic environment characterized by missing or thin markets, as typifies frontier regions in the nascent stages of economic development. We subsequently test the significance of these determinants using discrete choice analysis These modeling questions have particular relevance for informing carbon sequestration and global warming policy initiatives. Other on-going research conducted by the ecologists associated with the project focus on the species composition, abundance, structure, and re-growth of the different forests types in the region. In addition, litter and biomass studies have been completed which included carbon estimates for the different forest types, including forest re-growth on agricultural land, as function of fallow cycle dynamics. Fallow cycle dynamics are extremely important as the region is dominated by semi-subsistence agriculture with very little chemical inputs, so farmers depend on the fallow cycle to restore soil productivity. It will be these detailed data that will be used to calculate baseline carbon sequestration amounts.
热带森林砍伐对一系列与环境变化和经济发展研究有关的主题，包括全球变暖、土地退化、物种灭绝和可持续性问题，都具有重要意义。由于认识到森林砍伐的地点和模式往往同其规模一样重要，已促使越来越多的计量经济学研究将卫星数据和政府普查数据联系起来，目的是模拟毁林过程的空间方面。最初的研究集中于时间序列分析，而最近的工作已经开始开发利用土地使用的时间序列数据的模型。在本文中，我们利用近15年期间的3个日期的卫星数据来估计卫星像素处于森林或人为干扰状态的概率。我们的研究集中在跨越墨西哥南部坎佩切州和金塔纳罗奥州的农业边境的土地使用变化。该地区是美洲除亚马孙河流域外面积最大、历史最悠久的热带森林之一，已被认定为森林和生物多样性丧失的“热点地区”。在过去的30年里，这些森林一直承受着巨大的压力，因为1972年修建了一条高速公路，开放了移民的边界。这条公路是促进农业殖民化的更大发展努力的一部分，它促成了长期的土地改造，这一点已被专题地图绘图仪(TM)卫星图像捕捉到。我们通过装配一个空间数据库来捕捉这些景观动态，该空间数据库将1986-1997年的三张TM图像和其他空间环境和gis定位数据与政府人口普查和社会经济数据的数据连接起来。我们开发了一个简单的效用最大化森林砍伐决策模型。根据以前的研究，理论模型表明，在一个以缺少或市场稀少为特征的经济环境中，森林砍伐的许多可能决定因素是经济发展初期的前沿区域的典型特征。随后，我们使用离散选择分析测试了这些决定因素的重要性。这些建模问题对碳封存和全球变暖政策举措有特别的相关性。与该项目有关的生态学家正在进行的其他研究重点是该地区不同森林类型的物种组成、数量、结构和再生长。此外，还完成了凋落物和生物量研究，其中包括对不同森林类型的碳估计，包括作为休耕循环动力学功能的农业土地上的森林再生长。休耕周期动态非常重要，因为该地区主要是半自给农业，化学投入很少，因此农民依赖休耕周期来恢复土壤生产力。这些详细的数据将被用来计算基准的碳汇量。
京公网安备 11010802022788号
