search

ktoddbrown / PrecipOnSOC

Effects of precipitation frequency on SOC distribution
BSD 2-Clause "Simplified" License
0 stars 0 forks source link

Air temperature and precipitation #1

Open ktoddbrown opened 6 years ago

ktoddbrown commented 6 years ago

Here is a list of air temperature data sets:

https://www.esrl.noaa.gov/psd/data/gridded/tables/temperature.html

And the complimentary precipitation:

https://www.esrl.noaa.gov/psd/data/gridded/tables/precipitation.html

The CRU products appear to be at a coarser resolution so I'm leaning towards taking the following product: CPC Global Unified Gauge-Based Analysis of Daily Precipitation CPC Global Daily Temperature because they are on a relatively high resolution (0.5x0.5 degree), daily time step and from the same project.

@bpbond @vlbailey Thoughts?

bpbond commented 6 years ago

I don't have experience with CPC but it seems to fit the bill. Note that it's only 1979- though.

ktoddbrown commented 6 years ago

I would be a bit surprised to see anything earlier then that at a daily resolution. Or do I have too low an expectation? We could just go with 1980-1995 if we wanted to cut something that was representative of 'historical' conditions, not perfect but it would miss some of the accelerated climate change that has happened recently.

bpbond commented 6 years ago

It depends a bit on how this dataset will be used. As simply a climatology to compute MAT, MAP, etc., then fine. If we're trying to match timestamp-specific data (I don't think that's the case here) then we might be concerned about the limited timeframe. Anyway, I'd start with CPC but write analysis code agnostically so we don't have to change that if we swap in new climate data.

ktoddbrown commented 6 years ago

I have MAT pulled and calculated for nearest neighbor.

Any suggestions on precip metrics we want to use? I can think of the following 1) average number of events per year by day 2) MAP 3) length of average event by day

bpbond commented 6 years ago

I think we need to do a little research at this point. For example:

the Gini index (which describes how uniformly precipitation is distributed throughout a year) and the annual number of wet days https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2014GL062156

Also: https://www.nature.com/articles/ngeo2456

ktoddbrown commented 6 years ago

I pulled the Gini index into the analysis: 3454d24

It looks pretty simple and robust. No clue if we're going to see a trend in the analysis with it but I like the logic behind it (1 = punctuated rain events; 0 = even rain through the year). I calculated the Gini index annually from 1979 to 2017, removed years with >5% missing days, and calculated the mean annual Gini to use in the analysis.