Open calls in doctoral studies, the first period

Supervisors: Alexander Kmoch, Evelyn Uuemaa

Global ecological analyses face distortions from latitude-longitude data and map projections, impacting the estimation of environmental influence on water quality. In addition watershed delineation is severely hampered by the traditional projected global data. Machine learning and spatial modelling require data accessible through a standardized interface and representative to the Earth’s ellipsoidal surface. Discrete Global Grid Systems (DGGS) is emerging as a solution, that enables organizing spatio-temporal data at various resolutions. Hexagonal DGGS, with its multi-resolution approach, equal-area and equal neighbourhoods paradigm, facilitates consistent spatial analysis globally and provides access to covariates for machine-learning in a unified way. In addition, to improve machine-learning capabilites, the research will test applying General Adaptive Modelling (GAM) and Gaussian processes for spatial-temporal modelling on DGGS. The project aims 1) to apply methods from discrete differential geometry for more effective spatial data processing on the geometry meshes, particularly for hydrological modelling on DGGS hexagonal grids; 2) to develop methods for integrating diverse data into DGGS with minimal distortions, ensuring accurate spatial statistics and modelling. Overcoming these challenges is crucial for achieving consistent and accurate spatial water quality modelling on a global scale.