I am working with the CMIP5 (HADGEM2 AO) climate data to compute climatologies for present and future models. I wonder why pr (precipitation) and prc (convective precipitation) are separated and why the sum of both multiplied by 86400 s (as original data are given in kg/m2s) and aggregated over months on average overshoots actual monthly precipitation by 50 mm.
Hi Allan,
I have been using WRF model for my PhD research for couple of years. The WRF model also generates convective and non convective precipitations.
In general, a Numerical Weather Prediction (NWP) model with horizontal grid spacing (dx=dy) > 5km can not explicitly resolve cumulus clouds. Therefore, NWP model uses cumulus cloud parameterization to remove atmospheric instability. As a byproduct of convection schemes model produces precipitation. Hence it is termed as convective precipitation.
On the other hand, the microphysics parameterization scheme which simulates the process of hydrometeors phase change (e.g., water vapor into snow, ice into hail etc.) produces large scale or grid scale precipitation. This second kind of precipitation is also known as non-convective precipitation. To get the total precipitation in a certain time we need to add both the convective and non-convective components.
Unfortunately I am not able to answer rest of your questions as those are CMIP5 model specific, and I have never used CMIP5 data.
Thanks
Mostofa
Dear Allan,
Why do you use HadGEM2-AO? I recommend HadGEM2-ES. Maybe it is not appropriate for your domain. Some models overestimate and some of them underestimate. You need to choose the most reliable and accurate model. So it would be better if you could apply a multi-model ensemble approach.
Good luck,
Tufan
Hi Allan,
Pr and Prc is different. Pr may be the total precipitation, and Prc is just the precipitation from the convection, not including large-scale precipitation.
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series