GLM inflow values differ from lake.csv inflows— need help understanding why

12 views
Skip to first unread message

CG

unread,
Feb 3, 2026, 5:18:02 AMFeb 3
to Aquatic Ecosystem MOdelling Network (AEMON) discussion forum

Hello everyone,

I am new to GLM and I’m trying to understand why there is a noticeable gap between the in situ inflow values and the flow values computed in the GLM output file lake.csv (see attached graph).

I am running GLM with an hourly time step, using daily inflow and outflow values and hourly meteorological data (UTC format).

Is it normal to observe such a discrepancy between the input and simulated flow values? Could it be related to my configuration?

Here is an extract of my glm3.nml file:

&time
   timefmt = 2
   start = '2020-01-01 00:00:00'
   stop = '2024-12-31 00:00:00'
   dt = 3600.0
   !num_days = 2922
   timezone = 0 !tous les capteurs sont en utc
/

&inflow
   num_inflows = 3
   names_of_strms = 'creuse','petite_creuse','sedelle'
   strm_hf_angle = 60, 25, 52 !a changer ?
   strmbd_slope = 0.259, 0.259, 0.259 !pente du lit du cours d'eau -> a calculer encore
   strmbd_drag = 0.04, 0.04, 0.04 ! bottom roughness -> rate of entrainment as it enters the water body
   inflow_factor = 1, 1, 1
   inflow_fl = 'input_files/inflow_creuse_glm_v1.csv','input_files/inflow_petite_creuse_glm_v1.csv','input_files/inflow_sedelle_glm_v1.csv'
   inflow_varnum = 3
   inflow_vars = 'flow', 'temp', 'salt'
   coef_inf_entrain = 0
   time_fmt = 'YYYY-MM-DD'
/

&outflow
   num_outlet = 2
   outlet_type = 1, 1 ! outlet fixe 3= hauteur adaptee pour eviter faibles concentrations en o2
   outl_elvs = 186.1, 150 !
   bsn_len_outl = 5700, 760! verif pour le 30 pas des mesures exactes mesures dans mon espece d'ellipse
   bsn_wid_outl = 175, 257 ! verif
   outflow_fl = 'input_files/outflow_turbine_glm_v1.csv', 'input_files/outflow_reserve_glm_v1.csv'
   outflow_factor = 1, 1
   time_fmt = 'YYYY-MM-DD'
/


The time column in all inflow and outflow files is formatted as "YYYY-MM-DD".

If this mismatch is due to an error, do you have any idea where it might come from?

Thanks a lot for your help.

Best wishes,

Clara


 
Rplot01.png

Jordi Prats

unread,
Feb 6, 2026, 3:59:40 AMFeb 6
to Aquatic Ecosystem MOdelling Network (AEMON) discussion forum
Hi Clara,
if in your chart blue was GLM flow and red in situ flow, I'd say you're using the option "non_avg = .false.", which averages input data (flow and meteo) in consecutive days. If not, I don't know of any configuration option that would have the observed effect.
Best,
Jordi

El dia dimarts, 3 de febrer del 2026 a les 11:18:02 UTC+1, clara...@gmail.com va escriure:
Reply all
Reply to author
Forward
0 new messages