Dear John,
you may follow the FAO-56 (http://www.fao.org/docrep/X0490E/x0490e00.htm#Contents) descriptions and do the conversion manually in a spreadsheet - maybe first of all, to avoid duplication of work, just for a few months. The results you can compare with the weathermen outputs.
(I did this some time ago, and may still have the xls-spreadsheet, which I could send you.)

Regards,
Rolf

> -----Original Message-----
> From: DSSAT - Crop Models and Applications
> [mailto:[log in to unmask]] On Behalf Of John S. Felkner
> Sent: Monday, February 20, 2012 23:58
> To: [log in to unmask]
> Subject: low solar radiation values using Weatherman conversion from sunlight
> duration hours
>
> Dear Dssat Community:
>
> Using data on sunlight duration from a weather station in Thailand, we have
> converted sunlight duration hours to solar radiation in MJ/m2/day using the
> Weatherman sunlight duration/solar radiation conversion routine (when
> importing from the raw data import grid).  However, the resulting estimated
> solar radiation values in the Weatherman weather database that is created
> appear to be somewhat low:  none of the estimated solar radiation daily
> values exceed 20 MJ/m2/day, even during the summer months.
>
> We are wondering if anyone might have an idea why the estimated values
> seem
> somewhat low?  The input sunlight duration hours for a weather station in
> Surin Province, Thailand were obtained directly from the Thai
> Meteorological Department, and for 1998 range from 0 to 11.4, with a mean
> of 6.89 and a stdev of 2.72.  The estimated solar radiation values for the
> same days from Weatherman range from 4.6 to 18, with a mean of 12.5 and
> stdev of 3.02.  We specified "hours of bright sunshine" as the units in the
> Weatherman raw data import grid, and were careful to specify the correct
> latitude/longitude of the station as well as its elevation above sea level
> in meters.
>
> Thanks for any help or advice.  Is the Weatherman model for converting
> sunlight duration to solar radiation reliable?  Are there alternative
> methods for converting that anyone may be aware of?
>
> John Felkner
> University of Chicago