Fire Statistics Output Table Spec

[jdduh]

It turns out FGDB featureclass's attribute table won't allow field names that begin with a number. Please move the year numbers behind alphabetical characters. This modification will greatly improve the subsequent workflow to merge the tabular data with the geometry.

Just want to confirm the coding rule for missing data, zeros, and error codes in the output tables .

Blank: missing data
Zero: value of zero (data are available but with zeros or no burns)
-1: Errors?

[lbross]

Link to corrected design spec.

[lbross]

The coding rule for missing data, zeros, and error codes is different for nifc and mtbs data because vector and raster data are handled differently. Nifc will always have a feature class even if it doesn't have any rows. So the queries for that will return 0 for the years that don't have any data.
Mtbs data could be missing because there was no fire data that year or because the layer couldn't be clipped because there wasn't an imageservice for that year. The code looks at the collection of clipped files in the file .gdb rather than what is available on the server. For years that have no mtbs data missing data is represented as null because we just know it is missing, but not why.
The values for both always start as a -1 which means if any glitches prevent the process from completing, it will return a -1 error code.

[lbross]

Attached is sample of updated file. Please let me know if I got it right.
2019_annual_statistics.csv

[jdduh]

Sorry, I just commented on the column naming convention. Given that the field names might get truncated (e.g., when converting the table to a shapefile table), I prefer to keep the year information upfront. The current system I use is YY19_newfireno, for example. I know the output table headings can be different from what's in the attribute table of a featureclass, but I prefer that we use the same field names.

[lbross]

So the formula is YY + Last 2 digits of year + statistic description example: YY22_newfireno where the year is 2022. Is there any concern about representing the first two digits. Our current dataset still reaches back to 1995.

[jdduh]

Let's keep just the 2 digits year. If the tool is still being used in 2095, then I would be very happy! Besides, we only allow the the clipping of the most recent 30 years of data.

[lbross]

Attached a sample of an updated file. Please review. The selected end year for annual time period is not working in the version you have, but I fixed it.
1998_annual_statistics.csv

[lbross]

Always recalculate the area of the AOI. Store in session just in case. Remove aoishparea and aoishpunits from Constants class and all references. If the linear unit of the projection is not meters, write an error to the log that all statistics dependent on the aoi area may be incorrect due to projection mismatch. This is for mtbs and nifc statistics.