search

openwfm / WRF-SFIRE

A coupled weather-fire forecasting model built on top of Weather Research and Forecasting (WRF). This is the original https://github.com/openwfm/wrf-fire transitioned to a fork of WRF and selected as ifire=1. Graphic log at https://repo.or.cz/git-browser/by-commit.html?r=WRF-SFIRE.git
https://wiki.openwfm.org
Other
42 stars 12 forks source link

Transition test/em_fire ideal test cases #29

Open janmandel opened 4 years ago

janmandel commented 4 years ago

Transition remaining test/em_fire/* ideal problems ported from openwfm/wrf-fire to ifire=1, test, and fix.

janmandel commented 4 years ago

Using 80eb347933d6c5254819d21c22e1afb525276f6e Jul 30 in master: change fire_upwinding to sfire_upwinding with default 3 to avoid registry conflict with ifire=2 #22

janmandel commented 4 years ago

in branch develop at 63ff198963289230fe1

hill: working (does not crash and fire propagates, fire_log_interp=1, fire_use_windrf=0 - no Massman)

chem: working (does not crash and fire propagates)

rain1: made a copy of rain because rain is created dynamically; crash because of changes I had to make in the FMC PR wrf-model/WRF#792 changing names of FMC_GW01. Should work after I reconcile this conflict.

fireflux: seg fault in ideal.exe on writing wrfinput_d01 /glade/work/jmandel/WRF-SFIRE-debug and mpi reason: zsf dzdxf dzdyf not allocated (=have shape 1 1) in ideal.exe, even if they are present in wrfinput_d01 at full size (?). The fireflux case has 2 domains, and on domain 1, ifire=0 (no fire) , while sr_x=sr_y=1, resulting in out-of-bound writes into fire variables and memory corruption. Same in unmodified WRF4.2.1 /glade/work/jmandel/WRF-orig. Did this ideal case with 2 domains work in 3.4 before? Fixing this would require mods to dyn_em/module_initialize_fire.F to bypass all code referencing fire variables when ifire=0, which is mixed with code that needs to run in any case. Need multi-domain ideal anyway.

the rest not yet

janmandel commented 4 years ago

fireflux_med: namelist_quilt causes segfault, contents removed, then runs fine, in /glade/u/home/jmandel/w/WRF-SFIRE-mpi/test/em_fire/fireflux_med

janmandel commented 4 years ago

in /glade/p/univ/ucud0004/jmandel/WRF-SFIRE-mpi/test/em_fire fireflux_sfc 864 cores got to 409.79s sim run out of 12h max, ignition ok but max f_ros 0 fireflux_small: same start/end time as fireflux_med, but progress small, max f_ros 0.2831 for fireflux_med 2.3132

Fergui commented 1 year ago

Clean ideal tests have been pushed to https://github.com/openwfm/regression Some of them will be added to test/em_sfire