Diagnosis of GCM-RCM-driven rainfall patterns under changing climate through the robust selection of multi-model ensemble and sub-ensembles

Srishti Gaur (), Rajnish Singh (), Arnab Bandyopadhyay () and Rajendra Singh
Additional contact information
Srishti Gaur: Indian Institute of Technology
Rajnish Singh: Indian Institute of Technology
Arnab Bandyopadhyay: North Eastern Regional Institute of Science and Technology
Rajendra Singh: Indian Institute of Technology

Climatic Change, 2023, vol. 176, issue 2, No 12, 30 pages

Abstract: Abstract Understanding rainfall patterns is crucial for basin-wide risk management. The present study assesses rainfall patterns by smoothing their daily mean through Fourier fitting for the Subarnarekha basin of India. The adequate selection of the ensemble technique and corresponding best-performing regional climate models forced by global climate models (GCMs-RCMs) (sub-ensembles) has been carried out for projection of future rainfall patterns. The spatial performance metrics are used to select the GCMs-RCMs based on their ability to mimic the spatial patterns of the observed rainfall. The multi-model ensemble (MME) rainfall is generated by assimilating the simulated rainfall of selected GCMs-RCMs by employing statistical and machine learning (ML) techniques. Quantification of uncertainty in rainfall projections is performed through analysis of variance. Simple composite mean (SCM) statistical technique outperforms ML techniques. Optimum MME is obtained by combining 6-best performing sub-ensembles obtained from four spatial performance metrics (Fraction skill score, Goodman–Kruskal’s lambda Kling-Gupta efficiency, and spatial efficiency). The significant changes in rainfall patterns are obtained during 2010–2039, 2040–2069, and 2070–2099 with respect to the baseline period (1976–2005) as per Wilcoxon signed-rank test. An increase of 20–45% for RCP4.5 and 26–55% for RCP8.5 is obtained in peak mean rainfall per rainy day during future periods at both sub-basins. On the contrary, a decrease of 21–57% for RCP4.5 and 45–55% for RCP8.5 is obtained for trough (minimum) mean rainfall per rainy day during future periods. Our finding shows the possibility of early monsoon occurrences (8–30 days ahead) during future periods. Differences in projection between different choices of GCM-RCM models in the multi-model ensemble are the largest source of uncertainty (larger than differences between emission scenario or the effect of decadal variability). The overall finding of the study indicates that basin needs better preparedness to mitigate more erratic rainfall in future.

Keywords: ANOVA; Smoothing; Fourier fitting; Multi-model ensemble; Regional climate models; Spatial performance metrics (search for similar items in EconPapers)
Date: 2023
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10584-022-03475-z Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:climat:v:176:y:2023:i:2:d:10.1007_s10584-022-03475-z

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/10584

DOI: 10.1007/s10584-022-03475-z

Access Statistics for this article

Climatic Change is currently edited by M. Oppenheimer and G. Yohe

More articles in Climatic Change from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().