| Getting Started | Details | Problems using the program |
| General | Parameters | Runtime errors |
| Installation | Run option not working | |
| Upgrading | Technical | Why are my results not right |
| Using the user interface |
Provide a description of what you need. It is a lot easier for us to send an individual file if you simply need to check an equation. If you need the sources to completely recompile the program, please keep in mind that we do not often have requests of this sort and it is not a trivial matter to package up all the source code and support files needed to compile the program.
Usually if you terminate the process and/or restart the computer and run setup again this will fix the problem.
If you have correctly installed your WWW browser, the manual will be displayed using your browser when you request help in any of the screens or the Help Menu.
MS Word for Windows can also read HTML documents.
When I try to generate data having parameterized only from monthly summaries, ClimGen does not give solar radiation as an output.
Additionally, I see that CropSyst can estimate solar radiation from daily data one date at a time without having to perform a parameterization analysis.
Is it possible to get ClimGen to estimate solar radiation when parameterizing a location only with month summaries (I.e. there are no real historial daily data to estimate solar radition estimation parameters).
In the case of parameterization from real historial daily precipitation and temperature data, the solar radiation parameters are derived from a long term analysis of the data. During the generation phase, although ClimGen is producing precipitation and temperature values, it cannot do the long term analysis to derived the solar radiation parameters as each day of data is generated.
CropSyst can do this because it uses a slightly different method to estimate solar radiation where the solar radiation parameters are derived from latitude. The estimation method use by ClimGen (developed by Campbell and Donatelli) is better than the latitude derived method used by CropSyst. (Because of this, CropSyst users may get better results by using ClimGen to estimate solar radiation when real historial solar radiation is not available rather than letting CropSyst estimate the solar radiation.)
Once you have generated a set of data, you could try the parameter analysis using the generated precipitation and temperature data (rename the .GEN files to .DAT files and perform the 'site analysis'gain), but then the parameters for solar radiation estimation will be derived from non-real data and this may give unreasonable estimated solar radiation.
The problem is, for temperatures below zero, the calculation of saturation vapor pressure becomes asymptotically closer zero (See spreadsheet). The saturation vapor pressure for dew point temperatures is used to calculate the relative humidity so very slight changes in subfreezing temperatures (as in the random number used to generate temperature value) can result in apparently significant variation in relative humidity.
We did add an additional seasonal parameter to improve slightly the seasonal characterization of relative humidity. We also made the correction for vapor pressure over Ice in the Antoine equation for freezing conditions to account for super cooled conditions in accordance to the World Meteorological Organization recommendation. This also made a slight improvement. (See the second graph on the right in the spreadsheet)
But it does not appear that we will be able to make much more improvement of relative humidity during extreme freezing at this time.
Note that if your model is using relative humidity, it will most likely actually reverse the calculation to obtain the saturation vapor pressure, this is the number that is important for most models and it should be then correctly characterized from the relative humidity data produced by ClimGen, even though the relative humidity values themselves appear to be significantly off. Additionally these values will be during periods of the year not significant for most models.