Effects of outliers on remote sensing‐assisted forest biomass estimation: A case study from the United States national forest inventory

Abstract Large‐scale ecological sampling networks, such as national forest inventories (NFIs), collect in situ data to support biodiversity monitoring, management and planning, greenhouse gas reporting. Data harmonization aims link auxiliary remotely sensed field‐collected expand beyond field plots, but outliers that arise harmonization—questionable observations because their values differ substantially from the rest—are rarely addressed. In this paper, we review sources of commonly occurring outliers, including random chance (statistical outliers), definitions protocols set by temporal spatial mismatch between data. We illustrate different types effects they have on estimates above‐ground biomass population parameters using a case study 292 NFI plots paired with airborne laser scanning (ALS) Sentinel‐2 Sawyer County, Wisconsin, United States. Depending criteria used identify (sampling year, plot location error, nonresponse, presence zeros model residuals), many 53 Forest Inventory Analysis (18%) were identified potential single criterion 111 (38%) if all used. Inclusion or removal led substantial differences mean standard error estimate per unit area. The simple expansion estimator, which does not rely ALS other data, was more sensitive than model‐assisted approaches incorporated Including predictors showed minimal increases precision our relative models alone. Outliers causes can be pervasive workflows. Our serve note caution researchers practitioners inclusion unintended consequences parameter estimates. When inform large‐scale mapping, carbon markets, reporting environmental policy, it is necessary ensure proper use geospatial harmonization.