hydnam 2.0.0

Python implementation of NedborAfstromnings Model (NAM) lumped rainfall–runoff model

HydNAM

HydNAM is a Python implementation of the NedborAfstromnings Model (NAM), a lumped rainfall–runoff model.

This project is based on the NAM_Model created by hckaraman.

Installation

pip install hydnam

Getting Started

1. Prepare the Dataset

The dataset must contain the following properties: Date, Temperature, Precipitation, Evapotranspiration, and Discharge.

Date	Temperature	Discharge	Precipitation	Evapotranspiration
10/9/2016	15.4	0.25694	0	2.79
10/10/2016	14.4	0.25812	0	3.46
10/11/2016	14.9	0.30983	0	3.65
10/12/2016	16.1	0.31422	0	3.46
10/13/2016	20.1	0.30866	0	5.64
10/14/2016	13.9	0.30868	0	3.24
10/15/2016	11.1	0.31299	0	3.41
...	...	...	...	...

Ensure that the time intervals between dates are consistent (e.g., 24 hours) for accurate model performance.

2. Initialize the NAM Model

from datetime import datetime

from hydnam.dataset import Dataset
from hydnam.hydnam import HydNAM
from hydnam.parameters import Parameters

dataset = Dataset(
    timestamp=[
        datetime(2016, 10, 9),
        datetime(2016, 10, 10),
        datetime(2016, 10, 11),
    ],
    temperature=[15.4, 14.4, 14.9],
    precipitation=[0.0, 0.0, 0.0],
    evapotranspiration=[2.79, 3.46, 3.65],
    discharge=[0.25694, 0.25812, 0.30983]
)

params = Parameters(
    umax=0.01,
    lmax=0.01,
    cqof=0.01,
    ckif=200.0,
    ck12=10.0,
    tof=0.0,
    tif=0.0,
    tg=0.0,
    ckbf=500.0,
    csnow=0.0,
    snowtemp=0.0,
    qofmin=0.4,
    pmm=10.0,
    carea=0.9
)

nam = HydNAM(
    dataset=dataset,
    parameters=Parameters(),
    area=58.8,
    interval=24.0,
    start=None,
    end=None,
    spin_off=0.0,
    name="NAM"
)

print(f'Parameters: {nam.get_parameters().__dict__}')
print(f'Statistics: {nam.get_statistical_metrics().__dict__}')
print(nam.model_is_broken())

nam.plot_q(only_obs_and_sim=False).show()
nam.get_model_output().to_dataframe().to_csv("result.csv")

3. Optimize the Model

for m in ['SLSQP', 'L-BFGS-B']:
    nam.optimize(method=m)
    print(nam.get_parameters().__dict__)
    print(nam.get_statistical_metrics().__dict__)
    nam.plot_q(only_obs_and_sim=True).show()

The model will calculate and check which Parameters are optimal for the model and use it as the main Parameters for the model.

4. Customize Properties

nam.update_dataset(dataset, interval=24.0)
nam.update_parameters(Parameters())
nam.update_area(59.0)
nam.update_name("New NAM")
nam.update_spin_off(0.0)
nam.update_time_range(datetime(2016, 10, 9), datetime(2016, 10, 10))

5. Show Discharge

nam.plot_q(only_obs_and_sim=True).show()

6. Reload Model

When the model performs calculations incorrectly and throws an error, you should reset the properties to more reasonable ones and recalculate.

nam.reload()

License

This library is released under the MIT License.

Contact

If you have any questions or issues, please open an issue on GitHub or email us at [e-mail].