search

forc-db / Global_Productivity

Creative Commons Attribution 4.0 International
2 stars 0 forks source link

Polynomial models #11

Closed beckybanbury closed 5 years ago

beckybanbury commented 5 years ago

@teixeirak I've just run polynomial models on the variables and got some interesting results. The figures are saved here and the table with the p-values and r-squared values is here if you wanted to take a look. It looks like there is some level of decline in primary productivity at very high levels of precipitation, though it depends on which measure of precipitation is used. I'm also interested in the trends shown by ANPP_woody, as it seems to be responding to climate often in a completely different way to the other response variables. For example it seems to plateau at higher temperatures and lower latitudes, unlike the other measures of primary productivity, which continue to increase.

beckybanbury commented 5 years ago

@teixeirak It looks like polynomial models are better fit for the data; I've just run analysis on this using AIC values and the results are summarized here .

beckybanbury commented 5 years ago

@teixeirak I've also added new figures where the data has been scaled so it's easier to compare between response variables, saved here

beckybanbury commented 5 years ago

@teixeirak results for new climate variables are in this folder. I've added results for aridity, potential evapotranspiration and vapour pressure deficit. VPD data was downloaded from terraclimate. It looks like this was derived from the WorldClim data and a couple of other data sets. It looks like this isn't such a good predictor as vapour pressure, but interesting anyway.