Thanks for the feedback! I’ll check on that message about no fire data in old forecasts. It’s true, but it’s not good for it to stop wret.exe if the program is busy doing old comparisons (like for autolearn). Meanwhile, try unchecking the fire weather box in wret.exe (if in fact you’d checked it). Oh, and it IS latestfire.csv, not fire.csv!!
I’ll check on these things.
A little background on the FWI (Canadian) and FFDI 9Australian). The FWI is rather difficult to calculate directly, as there are all kinds of conditions, “if’s”, etc. However, I found a VERY extensive study (from Australia - don’t recall the link at the moment) comparing the two, not only in terms of utility and accuracy, but also in how to (sort of) convert one into the other. The FFDI is more straightforward to compute (though it still has pretty much the same inputs), so I actually am computing FFDI and coming up with FWI from that, using the relations found in the study, BUT I actually went further and used some additional data in the study, to alter the weightings of the different factors. So, with the same input, the sort of “fake” FWI should be very close to correct values. A caveat there, though, is that the study used Australian sites, so I wouldn’t be surprised if this “FWI” is a bit off from the real one … in Canada (it’s probably almost perfect in Australia). Actually, the toughest part was to incorporate something about fuel moisture. The FWI has very complicated rules for that. I approximated these (and a similar drought index in FFDI) as best I could, using recent precipitation and other climate data from WXSIM. That’s a pretty significant part of the calculation.
Thanks for the feedback!
Tom