Uribe-Rivera Et Al 2017 Dataset: High Resolution Bioclimatic Layers For Southwest Of South America For Three Recent Past Periods (1970, 1990 And 2010)

These files were generated as part of the article "Dispersal and extrapolation on the accuracy of temporal predictions from distribution models for the Darwin’s frog" (Uribe-Rivera et al. 2017; accepted in Ecological Applications) We used point data of meteorological stations between 34°-48°S and 70°-75°W, to generate new climatic surfaces for three recent past periods (1970; 1990; 2010). Meteorological data encompassed 293 weather stations, and were extracted from three databases: Dirección Meteorológica de Chile (DMC); Dirección General de Aguas de Chile (DGA); and the FAOClim-NET Agroclimatic database management system (FAO 2001), recording monthly records of mean daily minimum temperature, mean daily maximum temperature and total rainfall for 5-year periods (1965-1969 for 1970 climatic conditions; 1985-1989 for 1990 climatic conditions; and 2005-2009 for 2010 climatic conditions). For each period monthly mean values of each climatic variable were interpolated to generate surfaces using Anusplin v.4.4 (Hutchinson and Xu 2006), which applies the same algorithm used to derive the WorldClim bioclimatic surfaces (Hijmans et al. 2005). Interpolations were fitted following Pliscoff et al. (2014) at a ~1x1 Km resolution, with elevation as an independent variable using the GTOPO30 global digital elevation model (USGS, 1996). Because some weather stations do not have information for every month, we used the option of non-data of Anusplin. The quality of interpolations of climatic data was assessed calculating the Pearson correlation coefficient at the cell level between the monthly climatic values from the CRU-TS v3.10.01 Historic Climate Database for GIS (Climatic Research Unit - Time Series, 2012), and the monthly climatic values from the new climatic layers. Finally, surfaces of 19 bioclimatic variables were generated using the dismo package in R (Hijmans et al. 2014). All bioclimatic layers were uploaded in a single compressed ZIP file. Individual layers can be found inside it as georeferenced ASCII raster files, and nominated primarily by time period, and secundarily by the number of bioclimatic layer, following the worldclim nomenclature (http://www.worldclim.org/bioclim).本数据集相关文件源自论文《达尔文蛙分布模型时间预测精度的扩散与外推》（Dispersal and extrapolation on the accuracy of temporal predictions from distribution models for the Darwin’s frog），作者为Uribe-Rivera等，2017年，已被《Ecological Applications》接收。 本研究使用南纬34°至48°、西经70°至75°范围内的气象站点点位数据，针对三个近期时段（1970年、1990年、2010年）生成全新的气候表面数据。本次研究涉及的气象数据源自293个气象站点，采集自三个数据库：智利气象局（Dirección Meteorológica de Chile，DMC）、智利水利总局（Dirección General de Aguas de Chile，DGA）以及FAOClim-NET农业气候数据库管理系统（FAO，2001）。数据记录了对应5年时段的月均日最低气温、月均日最高气温与总降雨量：1970年气候条件对应1965-1969年时段，1990年对应1985-1989年时段，2010年对应2005-2009年时段。 针对每个时段，研究采用Anusplin v.4.4软件（Hutchinson与Xu，2006）对各气候变量的月均值进行空间插值以生成气候表面，该软件所采用的算法与生成WorldClim生物气候表面的算法一致（Hijmans等，2005）。 插值过程遵循Pliscoff等（2014）的方法，以约1×1公里的分辨率进行，同时引入GTOPO30全球数字高程模型（USGS，1996）中的高程数据作为自变量。 由于部分气象站点存在月度数据缺失的情况，本研究使用了Anusplin软件的无数据处理选项。 本研究通过计算单元格尺度下的皮尔逊相关系数（Pearson correlation coefficient）来评估气候数据插值的质量，对比数据分别来自CRU-TS v3.10.01 GIS用历史气候数据库（Climatic Research Unit - Time Series，2012）与本次生成的新气候图层的月气候值。 最后，本研究借助R语言的dismo包（Hijmans等，2014）生成了包含19个生物气候变量的气候表面数据。 所有生物气候图层均打包为单个压缩ZIP文件上传。文件内部以地理参考ASCII栅格文件格式存储各独立图层，命名规则遵循WorldClim的命名规范：以时段作为主命名依据，以生物气候图层编号作为次要命名依据，相关规范可参考http://www.worldclim.org/bioclim。