Drag sensors to move · click empty ground to add · scroll to zoom
Semivariogram: model γ(h) vs. empirical pair cloud

IoT Soil-Sensor Kriging: 2D Interpolation Map

A handful of wireless soil-moisture probes cannot cover a field pixel by pixel — so precision-agriculture networks lean on geostatistics to fill the gaps. This 2D simulator scatters a configurable number of sensors over a synthetic field, computes real ordinary kriging (spherical semivariogram, unbiasedness-constrained weights via a Lagrange multiplier) to build a continuous moisture surface from those sparse readings, and renders the kriged estimate, its per-cell estimation-variance map and the hidden ground truth as a top-down heatmap you can pan and zoom. Drag any sensor to a new spot, drop a new one, or delete one outright and watch the whole surface, its uncertainty map, the empirical-variogram cloud and the RMSE-vs-truth readout re-solve live — exactly the trade-off a real deployment has to budget for.