Hi all,
I have a database of monthly precipitation time series for several synoptic stations. I want to test the nonstationary of these time series. Two questions ariese here:
which tests can you propose for this purpose?
how one can konw that this nonstationary is in mean or variance?
Thank you very much for your help.
The following URL link is to a report that the Oregon Department of Transportation commissioned because the updates of precipitation maps were slow coming from the federal government.: https://www.oregon.gov/ODOT/Programs/ResearchDocuments/SPR656_Rainfall_Analysis_Final_Report_web.pdf
The updates of US precipitation frequency and intensity maps by NOAA (National Oceanic and Atmospheric Administration) for "Atlas 14" included extensive tests of stationarity. Given the careening band-wagon regarding the "Climate Crisis," NOAA dearly wanted to find/demonstrate non-stationarity in the precipitation data. I will leave it to the reader to go look at the results. Regardless, the various reports on the regional developments of the various parts of "Atlas 14" document the testing that was done for stationarity.
This might be helpful:
http://www.wmo.int/pages/prog/hwrp/chy/chy14/documents/ms/Stationarity_and_Nonstationarity.pdf
Hi dear Majid
You can use XLSTAT software for your research. There are several tests to the determination of datasets properties in the statistical analysis part of this software.
Good luck
I would recommend the Mann-Kendall test, as in this example:
Chapter Trend Detection in Annual Temperature and Precipitation Usin...
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series