Projections for four climate variables for national parks in RSA.
temporal frequency: daily
temporal coverage: 1950-2100
scenarios: ssp245 and ssp585, both contain historical and scenario experiment. Historical experiment covers 1950-2015, the scenario experiment 2016-2100.
variables: pr, tasmin, tasmax, huss (specific humidity)
units: pr: mm/day
tasmin: deg C
tasmax: deg C
huss: kg/kg

ensemble size: varies, depending on variable. 30 common models for ssp585, and 27 for ssp245

data format: csv
file contents: daily data with one value for each of the national parks


daily values for each of the parks are derived from gridded data as a mean value of the interpolated gridded field over that park's polygon
polygons used are in gis/QDS_SANParks.shp and basically represent the extent of a rectangular 0.5deg grid overlapping over the actual extent of the park.

IMPORTANT:
values are NOT BIAS-CORRECTED, i.e. they reflect GCM data as is. This is a significant limitation to how these data could be used. The constraints are as follows:
- data are likely to be biased, i.e. have a systematic offset compared to observations
- the bias is model-specific, with the consequence that differences between models in historical period reflect model bias, and in future period - a combination of model uncertainty and model bias
- for rainfall - properties such as number of rain days, consecutive dry days etc. are likely to be unrealistic