GloH2O

Menu

PBCOR

Precipitation Bias CORrection

The PBCOR dataset provides precip­i­tation bias-correction factors derived from streamflow observations. 

Overview

State-of-the-art gauge-based climatologies — such as WorldClim, CHPclim, and CHELSA — seriously underestimate precipitation over most major mountain ranges.

The Precip­i­tation Bias CORrection (PBCOR) dataset consists of global gap-free bias correction maps derived using streamflow obser­va­tions from 9372 stations worldwide. For each station, we inferred the “true” long-term precip­i­tation using a Budyko curve, which is an empirical equation relating long-term precip­i­tation, streamflow, and potential evapo­ration. We subse­quently calcu­lated long-term bias correction factors for WorldClim, CHPclim, and CHELSA, after which we used a random forest model to produce global gap-free bias correction maps for the clima­tologies. For further infor­mation, please see the following paper:

Beck, H. E., T. R. McVicar, M. Zambrano-Bigiarini, C. Alvarez-Garret, O. M. Baez-Villanueva, J. Sheffield, D. Karger, and E. F. Wood, 2020. Bias correction of global high-resolution precip­i­tation clima­tologies using streamflow obser­va­tions from 9372 catch­ments, Journal of Climate 33, 1299–1315, doi:10.1175/JCLI-D-19–0332.1.

Streamflow-based bias correction

Use the slider to compare the estimated and observed bias-correction factors for the WorldClim precip­i­tation clima­tology. Each point repre­sents a catchment centroid. 
Bias correction factor 

Data access

Freely use, adapt, and share the PBCOR dataset for non-commercial purposes, with attribution to Beck et al. (2020).

The latest version (1.0) of the PBCOR dataset, including a readme with infor­mation about the files, can be downloaded here.

The data are released under the Creative Commons Attri­bution-NonCom­mercial 4.0 Inter­na­tional License (CC BY-NC 4.0). This license permits use of the dataset for research and educa­tional purposes but prohibits commercial use. Proper attri­bution is required, and users must cite Beck et al. (2020) in any publi­ca­tions based on the dataset.

If the dataset plays a signif­icant role in your research, we kindly request the oppor­tunity to provide feedback on your results prior to publication.