VeriGeo: Runtime Validation and Repair for Projections, Areal Interpolation, Topology, and Spatial Statistics in Automated GIS

Automated GIS scripts can complete without runtime exceptions while still producing spatially invalid outputs. We present VeriGeo (NS-VGP 1.0), a Python runtime checking and repair layer for coordinate reference systems, areal interpolation, polygon topology, and spatial-statistics preconditions. We evaluate the implementation on a deterministic synthetic benchmark comprising 70 tasks (60 fault-injected tasks and 10 clean controls); the repository records 535 method executions across these tasks and ablations. The unverified GeoPandas baseline passed 0/60 fault-injected tasks and 10/10 controls (14.3% overall), with a 35.00% mean geodesic area error and a 169.56% mean population overcount on the benchmark joins. Under the benchmark's stated evaluator, VeriGeo passed 57/60 fault-injected tasks (95.0%) and 10/10 controls (67/70 = 95.7% overall; 95% Wilson CI [88.1%, 98.5%]), with a 25.44 ms mean execution latency and a 0.024% mean geodesic area error. Across the nine Gaussian-hotspot joins, per-zone population NMAE was 3.84% for VeriGeo versus 7.97% for area-weighted interpolation; including the unmasked water-barrier join, the ten-join means were 10.10% and 12.80%, respectively. Three tasks remain failures: the unmasked water barrier (MAUP-14), an ambiguously named integer-valued income field (MAUP-15), and a legitimate narrow canal removed by the sliver filter (TOPO-15). The statistics benchmark evaluates finite Moran-style outputs and selected spatial preconditions; it does not establish that GWR, LISA, or Getis-Ord estimators or inferential workflows have been validated.

Authors

Publication Details

Journal
EarthArXiv (California Digital Library)
Published
2026-11-16
DOI
https://doi.org/10.31223/x5s81p
Primary Topic
Geographic Information Systems Studies
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

VeriGeo: Runtime Validation and Repair for Projections, Areal Interpolation, Topology, and Spatial Statistics in Automated GIS

Prince
2 links
EarthArXiv (California Digital Library)
Geographic Information Systems Studies
preprint

VeriGeo: Runtime Validation and Repair for Projections, Areal Interpolation, Topology, and Spatial Statistics in Automated GIS

Prince
preprint en

Abstract

Automated GIS scripts can complete without runtime exceptions while still producing spatially invalid outputs. We present VeriGeo (NS-VGP 1.0), a Python runtime checking and repair layer for coordinate reference systems, areal interpolation, polygon topology, and spatial-statistics preconditions. We evaluate the implementation on a deterministic synthetic benchmark comprising 70 tasks (60 fault-injected tasks and 10 clean controls); the repository records 535 method executions across these tasks and ablations. The unverified GeoPandas baseline passed 0/60 fault-injected tasks and 10/10 controls (14.3% overall), with a 35.00% mean geodesic area error and a 169.56% mean population overcount on the benchmark joins. Under the benchmark's stated evaluator, VeriGeo passed 57/60 fault-injected tasks (95.0%) and 10/10 controls (67/70 = 95.7% overall; 95% Wilson CI [88.1%, 98.5%]), with a 25.44 ms mean execution latency and a 0.024% mean geodesic area error. Across the nine Gaussian-hotspot joins, per-zone population NMAE was 3.84% for VeriGeo versus 7.97% for area-weighted interpolation; including the unmasked water-barrier join, the ten-join means were 10.10% and 12.80%, respectively. Three tasks remain failures: the unmasked water barrier (MAUP-14), an ambiguously named integer-valued income field (MAUP-15), and a legitimate narrow canal removed by the sliver filter (TOPO-15). The statistics benchmark evaluates finite Moran-style outputs and selected spatial preconditions; it does not establish that GWR, LISA, or Getis-Ord estimators or inferential workflows have been validated.

EarthArXiv (California Digital Library)
Geographic Information Systems Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon