Hydrological parameters modelling in catchments based on a geographical database.

The water systems management plays a pivotal role in environmental conservation and disaster mitigation. As climate change intensifies, the ecological interactions of our ecosystems are modified, decreasing biodiversity and increasing extreme events. Therefore, accurate hydrological modelling tools are crucial for predicting rainfall-runoff processes. Hydrological processes, in general, are complex due to the interaction between multiple variables and spatial and time scales. Therefore, the development of hydrological models has evolved from simple models with few parameters to complex models aiming to model all notable processes within the study area. However, some researchers affirm that increasing the number of free parameters does not necessarily improve the model performance, and retaining only necessary data can ensure that the model’s components are positively represented. In this work, we show a set of geographical information system-based methodologies to set a limited optimal number of parameters to improve the hydrological modelisation.

To achieve our purpose, we collected terrain information, land use and soil properties data to model the water balance based on historical precipitation and gauging data. The same model was replicated in 47 small watersheds north of the Iberian Peninsula to ensure reliability. The rainfall and water flow data were downloaded from the automatic hydrology information system of the Ebro Water Confederation (SAIHEbro). We obtained a 15-minute rainfall and water flow time series, and each of them started at different years, continuing to current times up to a length of 27 years (more than 35,000 records per year).

As a result, we developed a database including the watershed limits, the most extended stream segment, rainfall and flow for each catchment. Furthermore, elevation, land use, soil classes, bulk density, clay, sand, and silt content (Hengl et al., 2017) at different depths were obtained. All data were transformed to a raster format to homogenise, and then their spatial resolution was harmonised to 2m for all spatial layers. The main shortcomings were found in matching the different spatial scales available in all the studied datasets. The lack of data or gaps in 2m DEM needed to be filled. Therefore, a nearest neighbour interpolation method combined with patching technique was performed by SAGA software and using 5m DEM as an input. Furthermore, differences in land use characterisation among regional and national datasets arose in some of the study catchments.

By processing these datasets, we obtained essential parameters for hydrological modelling. Altogether, the gathered information was useful to simulate the evolution of the water-related processes, paying particular attention to the relationships between precipitation, soil water content and land use.

Acknowledgements: The authors acknowledge the support of the Project “Fusión de modelos de base física y basados en datos para la modelización de fenómenos precipitación-flujo HYDER”, from Universidad Politécnica de Madrid (project number: TED2021-131520B-C21).

References

Hengl, T., De Jesus, J.M., Heuvelink, G.B.M., Gonzalez, M.R., Kilibarda, M., Blagotić, A., Shangguan, W., Wright, M.N., Geng, X., Bauer-Marschallinger, B., Guevara, M.A., Vargas, R., MacMillan, R.A., Batjes, N.H., Leenaars, J.G.B., Ribeiro, E., Wheeler, I., Mantel, S., Kempen, B., 2017. SoilGrids250m: Global gridded soil information based on machine learning, PLoS ONE. https://doi.org/10.1371/journal.pone.0169748