Weather stations assignation

[Viorel Chendes]

Dear ArcSWAT users,

I found here and also in some scientific papers some discussions related to the way in which climatic parameters (daily and monthly average) is interpolated / assigned at the subbasin level. 

Unfortunately, no one answer is concluded for my problem. It is still a mystery for me...

So, I try to use SWAT for a small basin (about 80 sq.km), where I have 3 pluviometric stations. For these points I have also temperatures (see Fig_1).

The location of the pcp points are as following:

ID,NAME,LAT,LONG,ELEVATION

1,pp_boroaia,22.3062,46.4767,360.5

2,pp_moneasa,22.2543,46.4601,288.8

3,pp_ranusa,22.2544,46.4237,229.5

After the input preparing and setting, I notice that only one weather station is taken into account for all subbasins. Nothing happens like average precipitation or Thiessen interpolation (as was revealed in a previous post).

This result can be verified in the SWAT Error Checker window - Hydrology tab, where water balance elements are highlighted. The precipitation amount showed in this fig is that of station number 1 - that is typically for mountain area (in the station 2 and 3, the precipitation is lower) - see Fig_2 file.

I documented more detailed this issue. I found that the tables with correspondence of weather stations and daily climatic parameters for each subbasin are located in the "project".mdb, the tables SubHmd, SubPcp, SubSlr, SubTmp, SubWgn, SubWnd.

I modified manually this files, but after SWAT Run, the result was the same: only one station is considered. Consequently, manually modification didn't work.

So, my effective questions are:

a) I have a question related to these files: why for my project, MinDist field have such huge values? For example, the lower value in this field is 3451662 (see Fig_3). I understand that this value means the minimum distance from watershed to weather station. I think is a bug here.

There is anybody who solved this problem?

b) It is possible that *txt file were weather stations with daily data to be wrong written? - see Fig_4.

I consider that the using way for climatic parameters (uncontrolled by users) under SWAT model is a high limitation.

Thank you in advance. Best regards,

Viorel Chendes

[Viorel Chendes]

Dear colleagues,

I detected my error. 

Unfortunately I made a big mistake which I saw only later. It is an example of small error but with big effects :)

So, because most GIS software provide geographic coordinates in "longitude, latitude" order, I fill-up weather_station*txt file in the same order. But in this file the coordinates must be written in LAT, LONG order (inverse). I did not pay attention to these values.

Consequently, my weather stations has been located by SWAT in Saudi Arabia instead of Romania :)). Thus, the [mindist] parameter is huge and is normal to be considered only one stations instead of all 3.

So, if you have a problem related to the omission of some weather stations, first step is to verify the correct order: LAT, LONG :)

Best regards,

Viorel Chendes

National Institute of Hydrology,

Romania