wknoben / marrmot Goto Github PK

https://github.com/wknoben/MARRMoT/blob/master/MARRMoT/Models/Flux%20files/saturation_7.m#L24

The integral operation leads to m_14_topmodel_7p_2s about 20 times slower than other models?
Any idea how to improve this?

The attachment is the consumed time for each model in my case.

Thanks and Regards,
Dongdong

In order to test the BMI it will be good to have a config file using ERA5 data.

Implement the ability to continue running the model using saved model outputs from the last run

In m_28_xinanjiang_12p_4s:

flux_e  = evap_21(lm,c,S1,Ep,delta_t);
flux_r  = saturation_14(a,b,S1,wmax,flux_pi);
flux_rs = saturation_2(S2,smax,ex,flux_r);
flux_ri = saturation_2(S2,smax,ex,S2*ki);
flux_rg = saturation_2(S2,smax,ex,S2*kg);
flux_qs = flux_rb + flux_rs;
flux_qi = interflow_5(ci,S3);
flux_qg = baseflow_1(cg,S4);

% stores ODEs
dS1 = flux_pi - flux_e  - flux_r;
dS2 = flux_r  - flux_rs - flux_ri - flux_rg;
dS3 = flux_ri - flux_qi;
dS4 = flux_rg - flux_qg;

Hence, I think JacobPattern should be

obj.JacobPattern  = [1,0,0,0;
                     1,1,0,0;
                     0,1,1,0;
                     0,1,0,1];

Other than

obj.JacobPattern  = [1,0,0,0;
                     1,1,0,0;
                     0,1,1,0;
                     0,0,1,1];
% the last row error

https://github.com/wknoben/MARRMoT/blob/master/MARRMoT/Models/Model%20files/m_28_xinanjiang_12p_4s.m#L34

on the python side, grids are not implemented

work in progress

Diagnosis:

• x2 (water exchange parameter) is determined by relative fullness of store S2 (R in original GR4J paper), which is controlled by x3(maximum routing store depth)
• Empty S2, no flow from x2; full S2, flow from x2 as x2's value
• Perrin (2001) says: ". In absolute value, F cannot be greater than x2: x2 represents the maximum quantity of water that can be added (or released) to (from) each model flow component when the routing store level equals x3."
• Theoretically, this works; if S2 never goes above x3, import based on x2 never goes above x2
• In practice, if S2 goes over x3, we can end up in situation where water import via x2 goes up (because S2 > x3), which makes S2 go up further, etc.
• At some point, an equilibrium will be reached where outflow from S2 through flux Qr is equal to import via x2

Why would S2 go over x3?

• Numerical inaccuracies, rounding, etc
• if x2 > x3

Possible solutions:

• Extra constraint in recharge_2 to ensure that returned flow is at most equal to "flux"
• Enforce a check that x2 cannot be greater than x3

start and end-time are date-vecs, but current time is scalar.

When I run a model in parallel mode, I get the following error "The range of the parfor-loop variable 'k' must evaluate to a row vector of consecutive increasing integers"
Please help!

In R, Xinanjiang model calibration only cost about 40s.
But in MARRMoT, the calibration is quite low efficiency. With the same forings, half hour passed, still no output.

Just wondering which part delays the computing efficiency? ODE or optimization function? Any idea?

Numerical solutions don't always return physically realistic values.

Suggestion:

• Remove constraints from individual flux functions
• Build in constraints as part of the solver, where we (1) check if a new solution dumps a store into the negative; (2) if so, reduce all relevant fluxes by a scaled amount.

Big change, so probably should be saved for a large development push.

Calling bmi methods via grcp4bmi and Docker is giving errors.

model.get_output_var_names() returns error with

Cell(['Fluxes leaving the model > flux_out', 'Internal fluxes > flux_in',
       'Storages (time = t) > S(t)', 'Water balance check > wb'],
      dtype=object) has type <class 'oct2py.io.Cell'>, but expected one of: (<class 'bytes'>, <class 'str'>)

model.get_value('flux_out') returns error with

Arrays with type {} cannot be transmitted through this GRPC channel

please add BMI description to the main readme file.

In L107, to consistent with FluxNames in L38, it should be flux_qseg flux_c flux_qhsrf flux_qhgw flux_qt

https://github.com/wknoben/MARRMoT/blob/master/MARRMoT/Models/Model%20files/m_13_hillslope_7p_2s.m#L38
https://github.com/wknoben/MARRMoT/blob/master/MARRMoT/Models/Model%20files/m_13_hillslope_7p_2s.m#L107

get_input_var_names and get_value use different names. And var flux_out contiains 2 vars, should be split into 2 vars Q and Ea