Adaptive geospatial data search engine for river basin observation systems

River basins hold special interest and challenge in the Earth observation and digital twin designs. Given the complexity of these natural systems, multi-modal and multi-spectral measurements with high resolutions are needed to describe their state. In the MultiGIS4Rivers project, we have three measurements sites at different river systems: the Baixada Maranhense (Brazil), the Umia Basin (Spain) and the Odra River (Poland, Czechia). Together, they amass hundreds of variables and extensive metadata. We will prototype a spatio-temporal search engine that is specifically catered for these domains. The engine operates both on observations from the sites and its metadata: metadata is checked for integrity, and the missing or corrupted data is identified in measurements dataset. The search engine adapts to the lack of data in a selected domain by doing heuristic analysis of identified temporal and spatial data gaps. The analysis is then coupled with interpolation and propagation models developed in the project, together providing complete model-optimized measurement datasets. The search engine is a part of the data management digital platform that facilitates scientists and stakeholders of the MultiGIS4Rivers. The platform is designed as a web server providing its users with observations and modelling products via API compliant with the OGC standards. In addition, it allows for the visualization of the search results, showing data availability and statistics. The proposed approach is expected to demonstrate how adaptive data discovery and gap-aware processing can support interoperable hydrologic digital infrastructures for river basin applications.