Thanks Leonardo!

I'm just collecting basic climate data for a interpretation of some literature data. I recieved the climate data already (via worldclim). My question focuses on the interpretation of the data. 

I need a simple parameter to express the seasonality of precipitation to devide the literature sites into semi-humid and humid tropical. The Coefficient of Variation of the monthly rainfall (over several years) seams to be a possible way to express seasonality in my opinion. But from which CV on a site is called semi-humid?

Let's talk about the SC-parameterization soon. I trying to get access to a really nice dataset of SC-field data here from Goiania. I will get back to you when I get it.

Best wishes,

Dennis

Dear Dennis,

interesting question. I am afraid I cannot give you a prompt answer. I had some experience with calculating CV of (gridded) climate data for the globe. I am not sure if one can attribute the cv of P directly to the classification to be semi-arid or semi-humid. First of all, there is no stationary in classification of those attributes. Some regions might got wetter, some drier, and this depends fully on the climate input (and the way you classify semi-humid and semi-arid regions). This is especial a problem for those "transition" zones (arid and humid zones would be a bit more stationary I guess). Secondly, the cv is strongly dependend on the availability, length and quality of input data. It is long time ago that I have worked with worldclim, and I do not remember if those data are quality checked. Thirdly, other physiographic effects might be have also a strong effect on CV, e.g. position in the relief and elevation of the gage. All in all, I think it is quite challenging to classify semi-arid and semi-humid regions based on CV of Precip only. Maybe you could do it using another way: UNEP: World Atlas of Desertification, Edward Arnold, Sevenoaks, 1992. (and some others) present a way to calculate aridity on basis of climate input. We also have done it to prepare input data for our WaterGAP model. If you are interested, I can assist (I have access to the recent climate forcings on a 0.5 deg resolution) with that. Maybe after such a classification, you could relate the CV of P to that "reference map".

Best,

Hannes

Dear Hannes,

thank you for your response to my question and your offer to help. I think you are right with your concerns. I already thought as well that a grouping by using the CV would be quite rough and limitated. However, I changed plan and used the worldclim data (see this publication concerning its data quality) 

Hijmans, R.J. et al., 2005: Very high resolution interpolated climate surfaces for global land area.

for a classification according to Köppen. Here, seasonality of precipitation is accounted via rainfall of the driest and wettest month.

Best,

Dennis