#******************************************************************************* # # R code and corresponding files to the manuscript: # Umirbekov, A., Peña-Guerrero, M. D., & Müller, D. (2022). # Regionalization of climate teleconnections across Central Asian mountains # improves the predictability of seasonal precipitation. # Environmental Research Letters, 17(5), 55002. https://doi.org/10.1088/1748-9326/ac6229 #******************************************************************************* File "CA_oscillations.rds" contains the R script of workflow for: 1) Calculation of long-term annual and seasonal climatologies for the delineated precipitation subregions; 2) Estimation of Spearman rank correlation between climate oscillation indices and the seasonal precipitation; 3) Field significance test via False Discovery Rate'; 4) Running SVR-based model for seasonal predictions. File "subregion_final.shp" is the shapefile of the deliniated precipitation subregions File "processed_indices.csv" contains indices of climate oscillations investigated in the study. All indices were rearranged with respect to their lead-lag times to the seasonal precipitation (totals for Feb-June preciputation). Index name convention: - all indices contain three character string at the beginning or middle that corresponds to abrreviation of climate ocillations. Eg. "soi" refers to SOI or Southern Oscillation Index" - "_lag_1" suffix denotes a lag with respect to the target year. E.g. "soi12_lag_1" refers to SOI index in December (i.e. December of the last year) - the file also contains 3-month averages of the monthly indices, in such cases they start with prefix "l_". E.g. "l_pdo11_lag_1" means 3-month average of the PDO (Pacific Decadal Oscillation), for Sep, Oct and Nov ("11" here corresponds to the last of the month) in the last year. File "SVR_parameters.xlsx" contains gamma and cost parameters for each subregion`s SVR model.