-The ETo calculator version 3.2 of FAO was used to computing PET.
- Water balance for each month was estimated by subtracting the mean PET from mean rainfall amount in each month. The estimated water balance was interpolated using the universal kriging method in ArcGIS 10.4 to determine their spatial pattern in the study area. -The trend in rainfall amount, PET estimated water balance, AI was determined from the non-parametric Mann-Kendall trend test statistics and regression statistics https://www.researchgate.net/publication/330557056_Rainfall_and_potential_evapotranspiration_patterns_and_their_effects_on_climatic_water_balance_in_the_Western_Lithoral_Hydrological_Zone_of_Nigeria [accessed Jul 03 2020].
SWAT can do well over many climatic zones. It has evapotranspiration methods of Hargreaves, Penman etc. It uses several climate-parameters: rainfall, temperature, humidity etc. So you choose the parameters you want depending on what you intend to achieve. It has a well-documented Manual and is also integrated with different GIS software like ArcMap, QGIS, Mapwindow or you can go for the source code if you prefer so.
There are wide variety of choices available to you. However, the choice would be governed by the characteristics of the area under consideration and the purpose of modelling. You can use Agricultural Catchments Research Unit (ACRU) for hydrological modelling, as well as for the assessment of impacts of land use changes and climate changes on hydrology in catchments with diverse climates and land uses. The other alternative would be to employ Lumped conceptual models if you wish to understand hydrological system dynamics in areas with limited data where the application of physically-based models is limited.
May I suggest a more hands on approach. Watering Queensland has been a proposal since 1939. Our suggestion to answer this hydrological problem includes some innovative engiineering shown at https://www.bosmin.com/HOME/BosminPlanB.pdf
From experience in African basins; Hydrologic modeling would vary due to purpose of work;
Your major outputs for particular work would be:-
a) Catchment Annual Yield,
b) Peak floods vs. Return Periods,
c) Design Flood analysis
d) Flow time series,
e) Catchment model of the site area
f) Hydrological analysis report, or
g) Sediment analysis.
You may need the aid of HEC1, HEC2, or HEC-HMS due to your purpose
- For Flood analysis (estimating sizing of the "spillway") it could be done with the help of the HEC-HMS model.
- HEC-HMS is good for the African water catchments (with seasonal rainfall of a continuous patterns), while "SCS" Method is preferable with large drainage areas, and when the Curve Number exceeds 50.
There are several hydrological models which you can use over the west Afrcia. Some of the best suited for ET, will be VIC, SWAT, IHECRAS. You may refer the following papers:
Article Multi-Model Approach to Assess the Dynamics of Hydrologic Co...
It depend on your other avaible data (land use/cover, time step of your data,...) and also on the main objectif of your modeling . you can use the free model: HecHMS, SWAT, GR, ... or Licenced Model.
I think integrated biosphere simulator (ibis) or Joint UK Land Environment Simulator (JULES) are good models and they take precipitation and potential evapotranspiration as the input datasets.
There are various hydrological models which were used over Africa and the selction of the model depends on the purpose of your work. In my opinion from experiance SWAT or HEC-HMS is better.