The multisite option does not reproduce exactly the statistics calibrated for some set of parameters.

[JavierDiezSierra]

For some set of parameters the STNSRP model simulates series with fewer wet days and therefore with higher vales of variance and lower mean precipitation.

Check:

        for nunu, rr in enumerate(stations):
            x_aux=XX[nunu]; y_aux=YY[nunu]
            cells_wet=list()
            for ccc in range(len(Intensity_cells)):
                if coordinates.lower()=='geographical':
                    dist_cell_grid = (haversine((y_aux, x_aux), (y_cells[ccc], x_cells[ccc]), unit=Unit.KILOMETERS))
                    if radio_cells[ccc]  > dist_cell_grid:
                        cells_wet.append(ccc)
                elif coordinates.lower()=='utm':
                    dist_cell_grid=(distance_f(x_aux, y_aux, x_cells[ccc], y_cells[ccc]))
                    if radio_cells[ccc]>dist_cell_grid/1000:
                        cells_wet.append(ccc)
                else:
                    raise Exception ('coordinates is wrongly defined, it should be geographical or utm')

[JavierDiezSierra]

To solve the problem with the mean precipitation and the fewer wet days I changed the following line:

old: Number_cells_per_storm = ipsilon - 1##Random poisson mean new: old: Number_cells_per_storm = ipsilon##Random poisson mean

[JavierDiezSierra]

@navass11, could you please run again the "STNSRP_test.iynb" with Seasonality_type: ''monthy" and 200 year of simulation.