I am trying to compare the rainfall data measured at several met station locations with the netCDF output from WRF. I can create tslist for exact locations matching the met-mast lat/lon and generate the single point TS files. However, I am not convinced this may be the optimal way for comparison considering the assumptions and approximations in the cumulus and microphysics models and there can be deviations of few cells. Any suggestions or references will be appreciated.
Thanks.
If I understand well your question:
Given that the station precipitation is your "true" value, probably you need to use Ranked Probability score.
Wilks,D.(2011)Statisticalmethodsintheatmosphericsciences. Elsevier, 3edn.
Harish, What's the purpose of your comparison? Depending on your purpose, there are likely different answers. For example:
- estimating the quality of WRF model outputs
- incorporating WRF into multi-sensor precipitation analysis
- assimilating rainfall observations into WRF models runs
- etc
And don't forget that rain gage observations are observations at a point whereas atmospheric model outputs (as well as satellite and radar based rainfall estimates are are "representative" of possible rainfall over an area. There can be quite a difference between the two.
Thanks Geoffrey.
We are interested in analysing the performance of the Cumulus and Microphysics schemes for predicting the flash flood occurrence in Tropics. WRF precipitation output will be compared with rain gage observations and then trained using a machine learning model to possibly remove any systemic bias in the forecast.
You might take a look at the statistics used for 4D Data Assimilation at ECMWF and NWS. A few science to operations researches worth looking at:
- Dong-Jun (D. J.) Seo
- James. D. Brown
- Yuqiong Liu
- Julie Demargne
and:
- The Schaake Shuffle: A Method for Reconstructing Space–Time Variability in Forecasted Precipitation and Temperature Fields
- Demargne, J., Wu, L., Seo, D.-J., and Schaake, J., Experimental hydrometeorological and hydrologic ensemble forecasts and their verification in the U.S. National Weather Service, IAHS Publication. Quantification and Reduction of Predictive Uncertainty for Sustainable Water Resources Management (Proceedings of Symposium HS2004 at IUGG2007, Perugia, July 2007)., 313, , 177-187, January 2007
- Demargne, J., Mullusky, M., Werner, K., Adams, T., Lindsey, S., Schwein, N., Marosi, W. and Welles, E., Application of Forecast Verification Science to Operational River Forecasting in the U.S. National Weather Service, Bulletin of the American Meteorological Society, 90, 6, 779-784, January 2009
- Schaake, J., Demargne, J., Hartman, R., Mullusky, M., Welles, E., Wu, L., Herr, H., Fan, X. and Seo, D.-J., Precipitation and temperature ensemble forecasts from single-value forecasts, Hydrology and Earth Systems Sciences Discussions, 4, , 655-717, January 2007
- Julie Demargne, James Brown, Yuqiong Liu, Dong-Jun Seo, Limin Wu,1 Zoltan Toth, Diagnostic verification of hydrometeorological and hydrologic ensembles, Atmospheric Science Letters by the Royal Meteorological Society, , , , April 2010
- Yuqiong Liu, James Brown, Julie Demargne, Dong-Jun Seo, A wavelet-based approach to assessing timing errors in hydrologic predictions, Journal of Hydrology, 397, 3-4, 210-224, February 2011
- Julie Demargne, Limin Wu, Satish K. Regonda, James D. Brown, Haksksu Lee, Minxue He, Dong-Jun Seo, Robert Hartman, Henry D. Herr, Mark Fresch, John Schaake, and Yuejian Zhu, The Science Of NOAA’S Operational Hydrologic Ensemble Forecast Service, Bulletin of the American Meteorological Society, , , 1-21, January 2014
- Badges
- Science topic
- Similar topics
- Mathematics