Fig 1.
The interface includes functionalities such as searching the data catalog, drawing or uploading regions of interest, creating spatio-temporal domains, loading and visualizing layers from GEE, creating custom variable expressions, and downloading the resulting cluster or similarity maps.
Fig 2.
A user uploads a region of interest, sets the number of clusters k, the period is defined, selects relevant geospatial variables, and finally runs the clustering algorithm. The resulting clusters group areas with similar geospatial characteristics.
Fig 3.
Similarity search workflow in Sims.
A user defines the search and reference regions, sets the temporal period, optionally configures land cover masking and distance metrics, selects relevant variables from GEE, and runs the similarity search. The heat map highlights areas sharing similar features with the reference region.
Table 1.
Features included in each domain for clustering the simulated maize yields in Rwanda.
Fig 4.
Simulated maize yield (kg/ha) distributions across clusters in Rwanda.
Testing different numbers of clusters with Agronomy features revealed highly significant differences () between most cluster pairs at k = 5, except for some pairs (e.g., clusters 2 and 5).