Performance evaluation of regional climate models (RCMs) in determining precipitation characteristics for Gothenburg, Sweden
(2014) In Hydrology Research 45(4-5). p.703-714- Abstract
- Regional climate models (RCMs) are used for forecasting future climate including precipitation characteristics. Performances of five different RCMs for predicting the precipitation characteristics for Gothenburg, Sweden were evaluated against the daily observed precipitation over the period 1961 to 2009. Statistical analysis was done on annual, monthly, multi-daily, and daily data. The statistical techniques used include principal component analysis (PCA), comparison of annual maximum, frequency of exceedances determined from Poisson distribution, comparison of frequency distributions, and Mann-Kendall technique for investigating trend over time. Inter-annual variability and autocorrelation between years were also investigated. The results... (More)
- Regional climate models (RCMs) are used for forecasting future climate including precipitation characteristics. Performances of five different RCMs for predicting the precipitation characteristics for Gothenburg, Sweden were evaluated against the daily observed precipitation over the period 1961 to 2009. Statistical analysis was done on annual, monthly, multi-daily, and daily data. The statistical techniques used include principal component analysis (PCA), comparison of annual maximum, frequency of exceedances determined from Poisson distribution, comparison of frequency distributions, and Mann-Kendall technique for investigating trend over time. Inter-annual variability and autocorrelation between years were also investigated. The results obtained point towards the usefulness of these high-resolution RCMs. It was observed that all the models give the annual maximum precipitation within 3 mm of the observed data. As for the observed series, no trends were found for monthly or seasonal data. The number of exceedances above threshold accepted Poisson distribution hypothesis with the mean exceedances from RCM-PROMES being very close to the mean exceedances from the observed data. PCA also indicated that PROMES came closest to explaining the observed data. The presented statistical methods can be used for bias correction of raw RCM data in future studies. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/4728918
- author
- Rana, Arun LU ; Madan, Shilpy and Bengtsson, Lars LU
- organization
- publishing date
- 2014
- type
- Contribution to journal
- publication status
- published
- subject
- keywords
- climate change, daily precipitation, extreme precipitation, Gothenburg, regional climate models (RCMs), statistical techniques
- in
- Hydrology Research
- volume
- 45
- issue
- 4-5
- pages
- 703 - 714
- publisher
- IWA Publishing
- external identifiers
-
- wos:000341061400016
- scopus:84906915844
- ISSN
- 1998-9563
- DOI
- 10.2166/nh.2013.160
- language
- English
- LU publication?
- yes
- id
- 9a70fbc9-30d7-46a9-8e07-f125d9251af0 (old id 4728918)
- date added to LUP
- 2016-04-01 14:09:41
- date last changed
- 2022-03-21 22:28:09
@article{9a70fbc9-30d7-46a9-8e07-f125d9251af0,
abstract = {{Regional climate models (RCMs) are used for forecasting future climate including precipitation characteristics. Performances of five different RCMs for predicting the precipitation characteristics for Gothenburg, Sweden were evaluated against the daily observed precipitation over the period 1961 to 2009. Statistical analysis was done on annual, monthly, multi-daily, and daily data. The statistical techniques used include principal component analysis (PCA), comparison of annual maximum, frequency of exceedances determined from Poisson distribution, comparison of frequency distributions, and Mann-Kendall technique for investigating trend over time. Inter-annual variability and autocorrelation between years were also investigated. The results obtained point towards the usefulness of these high-resolution RCMs. It was observed that all the models give the annual maximum precipitation within 3 mm of the observed data. As for the observed series, no trends were found for monthly or seasonal data. The number of exceedances above threshold accepted Poisson distribution hypothesis with the mean exceedances from RCM-PROMES being very close to the mean exceedances from the observed data. PCA also indicated that PROMES came closest to explaining the observed data. The presented statistical methods can be used for bias correction of raw RCM data in future studies.}},
author = {{Rana, Arun and Madan, Shilpy and Bengtsson, Lars}},
issn = {{1998-9563}},
keywords = {{climate change; daily precipitation; extreme precipitation; Gothenburg; regional climate models (RCMs); statistical techniques}},
language = {{eng}},
number = {{4-5}},
pages = {{703--714}},
publisher = {{IWA Publishing}},
series = {{Hydrology Research}},
title = {{Performance evaluation of regional climate models (RCMs) in determining precipitation characteristics for Gothenburg, Sweden}},
url = {{http://dx.doi.org/10.2166/nh.2013.160}},
doi = {{10.2166/nh.2013.160}},
volume = {{45}},
year = {{2014}},
}