Oops ... first real calculation error in openGUTS

General announcements related to GUTS. Feel free to advertise new papers, courses, software, etc., as long as they are GUTS-related.

Moderators: Tjalling, Roman

Post Reply
User avatar
Tjalling
Site Admin
Posts: 187
Joined: Wed Sep 30, 2015 9:57 am
Location: The Netherlands
Contact:

Oops ... first real calculation error in openGUTS

Post by Tjalling »

The IT calculations in openGUTS use a shortcut that, I just found out, does not work in all cases. This error will affect IT calibrations for time varying exposure, where the exposure scenario includes episodes of linear decrease over time (for constant exposure and block pulses, it will be fine). It will also affect IT LPx predictions when the exposure scenario is specified with a crude resolution (the hourly resolution of FOCUS profiles is enough to prevent large errors; preliminary testing has not found an error >0.5%). This error will be repaired in the next update of the Matlab version, to v. 1.1, which will be soon. A document explaining this error in more detail will be posted on the openGUTS website even sooner.

Edit: on 26 May 2020, I released a new version 1.1 of the Matlab version of openGUTS that fixes this error. The detailed error document was placed on the website on the 19th. Further testing showed that it is difficult to come up with a reasonable data set for which this error has really large consequences (i.e., a data set for which the parameters can be reasonably identified).
Post Reply