Reconciling high resolution climate datasets using KrigR
Reconciling high resolution climate datasets using KrigR
复制标题
DOI：
10.1088/1748-9326/ac39bf
复制
发表时间：
2021-12-01
影响因子：
6.7
通讯作者：
Kusch, Erik
中科院分区：
文献类型：
--
作者：
Davy, Richard;Kusch, Erik
对气候变化及其后果感兴趣的广大研究人员越来越需要高空间和时间分辨率的气候数据。目前，全球气候模型和再分析数据集的空间分辨率（最好分别为0.25度和0.1度左右）与这些数据集的许多终端用户所需的分辨率之间存在很大的不匹配，这些数据集通常在30弧秒（类似于900米）的尺度上。由于需要提高气候数据集的空间分辨率，一些研究小组在统计上缩小了观测或再分析数据集的各种组合。然而，所使用的各种缩小尺度的方法和输入使得难以调和这些高分辨率数据集之间的差异。在这里，我们利用KrigR R-包，以统计上缩小世界领先的ERA 5（土地）再分析数据使用克里金。我们表明，克里金法可以准确地恢复气候数据的空间异质性，因为与协变量有很强的关系;通过保留与统计降尺度相关的不确定性，可以调查和解释高分辨率气候数据的置信度; KrigR提供的统计不确定性可以解释广泛使用的高分辨率气候数据集之间的差异（CHELSA、TerraClimate和WorldClim 2），取决于变量、时间尺度和区域。这表明了使用KrigR生成定制的高空间和/或时间分辨率气候数据的优势。
There is an increasing need for high spatial and temporal resolution climate data for the wide community of researchers interested in climate change and its consequences. Currently, there is a large mismatch between the spatial resolutions of global climate model and reanalysis datasets (at best around 0.25 degrees and 0.1 degrees respectively) and the resolutions needed by many end-users of these datasets, which are typically on the scale of 30 arcseconds (similar to 900 m). This need for improved spatial resolution in climate datasets has motivated several groups to statistically downscale various combinations of observational or reanalysis datasets. However, the variety of downscaling methods and inputs used makes it difficult to reconcile the resultant differences between these high-resolution datasets. Here we make use of the KrigR R-package to statistically downscale the world-leading ERA5(-Land) reanalysis data using kriging. We show that kriging can accurately recover spatial heterogeneity of climate data given strong relationships with co-variates; that by preserving the uncertainty associated with the statistical downscaling, one can investigate and account for confidence in high-resolution climate data; and that the statistical uncertainty provided by KrigR can explain much of the difference between widely used high resolution climate datasets (CHELSA, TerraClimate, and WorldClim2) depending on variable, timescale, and region. This demonstrates the advantages of using KrigR to generate customized high spatial and/or temporal resolution climate data.