Abstract. Climate Hazards Group InfraRed Precipitation with Stations (CHIRPS) was productedby blending Satellite-only Climate Hazards Group InfraRed Precipitation (CHIRP) with Stasion observations data. The blending process was aimed to reduce bias of CHIRP. However, Biases of CHIRPS on statistical moment and quantil values were high during wet season over Java Island. This paper presented a bias correction scheme to adjust statistical moment of CHIRP using observation precipitation data. The scheme combined Genetic Algorithm and Nonlinear Power Transformation, the results was evaluated based on different season and different elevation level. The experiment results revealed that the scheme robustly reduced bias on variance around 100% reduction and leaded to reduction of first, and second quantile biases. However, bias on third quantile only reduced during dry months. Based on different level of elevation, the performance of bias correction process is only significantly different on skewness indicators.
Nonlinear (NL) method is the most effective bias correction method for correcting statistical bias when observation precipitation data can not be approximated using gamma distribution. Since NL method only adjusts mean and variance, it does not perform well in handling bias on quantile values. This paperpresents a scheme of NL method with additional condition aiming to mitigate bias on quantile values. Non-dominated Sorting Genetic Algorithm II (NSGA-II) was applied to estimate parameter of NL method. Furthermore, to investigate suitability of application of NSGA-II, we performed Single Objective Genetic Algorithm (SOGA) as a comparison. The experiment results revealed NSGA-II was suitable when solution of SOGA produced low fitness. Application of NSGA-II could minimize impact of daily bias correction on monthly precipitation. The proposed scheme successfully reduced biases on mean, variance, first and second quantile However, biases on third and fourth moment could not be handled robustly while biases on third quantile only reduced during dry months.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.