I have a plan to compare the performances of statistically downscaled data and the regional climate model data (CORDEX). Statistically we can downscale in point scale but the RCMs are of 50 km grid. Do you think it is wise and logically fit to compare these two different products?
Hi Jeeban,
Yes it is possible to compare these two models because there is a strong correlation between them. In fact it is possible to calculate a linear relation between them using linear regression. See for instance:
https://www.researchgate.net/publication/239823315_WIND_HAZARD_USING_HIGH_RESOLUTION_CLIMATE_MODELS
Hope this helps,
Augusto
Article WIND HAZARD USING HIGH RESOLUTION CLIMATE MODELS
Hello Jeeban,
Yes, you can proceed with that, in fact, there are many literature performed such studies. RCMs unlike GCMs also suffer from systematic biases and have trouble capturing certain climatic elements (e.g. convective precipitation) over complex terrains. I would prefer both of them to be downscaled and then compared. Because, if you downscale an RCM you can address the improvements as compared to the raw data. use of downscaled regional climate model data is more effective in hydrological studies then the use of conventional downscaled GCM or raw RCM data.
Cheers :)
Hello,
I think it is not a fair comparison, because of the scale mismatch: The RCM grid cell represents the average over a relatively large area (in CORDEX mostly 50x50km). Additionally, the SDS has inherently done a bias correction with the observation stations.
It would be okay to do a post-processing bias correction of the RCM grid cells. Then, you can compare both, the SDS and the bias corrected RCM with the observation stations.
Another comment: Please do not just compare the values, please have a look at the Copulas, i.e. their dependence structure in the rank space, between downscaled and observed values. This gives you deeper insights, hopefully ...
e.g.:
Laux et al., 2011 (in HESS)
DOI: 10.5194/hess-15-2401-2011
Hi everyone,
I am also trying to perform similar comparison as Jeeban has mentioned; i.e. statistically downscaled AOGCM vs. RCM (CORDEX). I have already statistically downscaled my GCM data on 25x25 km resolution. It has a spatial mismatch with RCM which is on 50x50 km resolution. So, how can I downscale my CORDEX RCM to 0.25x0.25 degrees? Kindly advice.
Regards,
Raazia
Hi Raazia,
Spatial interpolation of climatic variables is not a trivial issue, the best interpolation techniques take into account a large number of variables like altitude, location, vegetation and information about climatic conditions in the region of interest, see for instance:
Article Spatial Interpolation of McArthur's Forest Fire Danger Index...
Hope it helps,
Augusto
- Badges
- Science topic
- Similar topics
- Geoscience
- Atmospheric Sciences
- Atmosphere