I am using RCMs to project the precipitation and temperature in Rawal watershed, Islamabad, Pakistan. I am using quantile mapping for bias correction. I want to know the statistical parameters which should be calculated to evaluate the performance of quantile mapping?
Mr Mubashir_Majeed
I have read some similar studies, that talk about the analysis of extreme events with rainfall evaluation of regional climate models, which I see you will benefit . for example Andrew Halmstad, Mohammad Reza Najafi and Hamid Moradkhani 2012 were did a study in Analysis of precipitation extremes with the assessment of regional climate models over the Willamette River Basin, USA . Thy evaluated the performance of quantile mapping using RCMs .
Also a modern study was done by agroup of scientists in 2015 M. A. Sunyer1, Y. Hundecha2, D. Lawrence3, H. Madsen4, P. Willems5,6, M. Martinkova7, K. Vormoor8, G. Bürger8, M. Hanel7,9, J. Kriaucˇiu ̄niene ̇10, A. Loukas11, M. Osuch12, and I. Yücel13.
They did a kind of study in Europe , they addressed with
Inter-comparison of statistical downscaling methods for projection of extreme precipitation in Europe.
https://www.narccap.ucar.edu/doc/pubs/halmstad-hyp9376.pdf
http://www.hydrol-earth-syst-sci.net/19/1827/2015/hess-19-1827-2015.pdf
per definition, the qm approach corrects for the statistics of the distribution. Thus, it is clear that the cdf of the bias corrected time series are improved, but you should look at parameters such as the Pearson r, RMSE, MAE, etc. Depending on your interests, it could be also relevant to check these parameters for different parts of the distribution (e.g. extreme values).
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis