Evaluation of Altimetric Accuracy in the Integration of GNSS Data withGlobal and Regional Digital Terrain Models in the Context of RuralProperty Georeferencing
Palavras-chave:
Digital Terrain Model, Orbital Sensors, GNSS, Data integration, AccuracyResumo
Digital Terrain Models (DTMs) are important in representing Earth’s surface relief. Recent updates of Brazil’s Technical Standard for Georeferencing of Rural Properties have made altimetric information a mandatory requirement for cadastral boundary delineation. Leveraging this regulatory advancement, the present study examines how the integration of GNSS survey data with both global elevation models and orbital radar sensors can enhance vertical accuracy and spatial coherence of DTMs in two Paraná municipalities (Reserva do Iguaçu and Guarapuava). This study aimed to evaluate the altimetric accuracy of integrating GNSS data with the Shuttle Radar Topography Mission (SRTM) and models derived from Sentinel-1 and ALOS PALSAR InSAR imagery. GNSS+SRTM integrated models were produced in QGIS, while Sentinel-1 and ALOS PALSAR were conducted in ESA’s SNAP environment. Model accuracy was quantified through Mean Absolute Error (MAE), Standard Deviation (SD), and Root Mean Square Error (RMSE), with independent GNSS points reserved as checkpoints. Spatial diagnostics comprised error density curves, boxplots, and Moran’s Index to assess residual clustering. Results revealed that GNSS+SRTM_IDW exhibited the poorest altimetric accuracy (RMSE up to 13.84 m; SD up to 6.88 m), reflecting high dispersion and systematic bias. In contrast, GNSS+SRTM_TIN consistently achieved the lowest errors (RMSE 1.99–2.88 m; MAE 1.26–1.91 m) and the most compact, symmetric error distributions. Among radarbased models, ALOS PALSAR outperformed Sentinel-1, with RMSE between 2.14–3.14 m versus 3.36–4.93 m, and demonstrated stronger spatial coherence. These findings underscore that TIN-based integration of GNSS and SRTM data - as well as ALOS PALSAR InSAR - offer the highest combination of accuracy and spatial reliability for altimetric applications. The study highlights the necessity of coupling global error metrics with spatial autocorrelation analyses to fully characterize DTM quality.
