The fact that many literature reports that the observation network in Africa has been deteriorating means that the qualities of the gridded products created from these observations have also been deteriorating. My question is what is the best way to use these products in the study of climate change and variability?
Dear Kwame,
What type of network are you referring to in this question?
However, you have thought about giving a statistical approach to the problem with current and old data, looking for variations and incorporating them into the problem.
Eduardo.
Thank you Diosdado Eduardo Marín Dorta, I am referring to the station datasets (both precipitation and temperature). Having in mind that the quality of the gridded products ultimately depends on the number, spatial distribution, and quality of the observations used.
My dilemma is if these observed in-situ datasets are deteriorating how best can we trust the gridded products? I stand to be corrected but I know a number of gridded datasets are generated by interpolating station measurements onto regular grids.
For instance, if you want to use them for application such as trend analysis, looking at the situation that the year-to-year variations may be influenced as much by the inhomogeneity of the time series. However, as you suggested when subjected to rigorous statistical approach, we can make headway.
Dear Collins Kwame Oduro I think your concern is valid. To check it, I would recommend the following procedure. Try to remove real stations one by one from the datasets. Use the remaining stations and interpolation scheme to get the temperature and precipitation on the location of the removed station. If the interpolated and real value have a large disparity, I think your concern is confirmed.
Best regards
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation