Integrating citizen-science generated data with the JULES land-surface model for hydrological analysis in the tropical Andes

Land surface model has been increasingly used for hydrological assessment because of their state-of-the-art representation of physical processes and versatility. In this study, the tropical Andean hydrology has been explored by the use of the JULES land surface model. The physically-based model has the advantage to map the modeller's knowledge about the hydrological impacts of land-use and land-cover change (LUCC) into physically meaningful parameters. However, the data availability could hinder the development of hydrological model, given the intensive data requirement to parameterise the complex environment with its complexity of meteorological and geographical conditions combined with extremely heterogeneous land-use. This major weakness of land surface models has been effectively reinforced due to the innovation in sensing technology and the expansion of monitoring network. We explore the use of high temporal resolution data collection on streamflow, precipitation, and several weather variables by a grassroots initiative (called iMHEA). In which, the JULES model has been initialised with using multiple sources of global data, and then been further calibrated for each land-use represented in the iMHEA dataset. The calibrated parameters have partitioned well on the surface and subsurface flows, hence, give more precise estimation on water balance. Wide range of hydrological responses under different LUCC has also been characterised by integrating citizen science data and land surface model. The constrained parameters using such distributed citizen-science generated streamflow data could further strengthen the JULES model in regional scale. The effectiveness of watershed interventions could be systematically assessed, which could be leveraged by catchment managers accordingly.