From Satellite Availability to Financial Information: An Auditable Sentinel-1B Benchmark

This paper examines whether satellite observation availability translates into useful economic information, using the December 2021 interruption of Sentinel-1B as a case study. The analysis links 176 mining sites and 38 listed issuers with public satellite acquisition records, administrative production data, radar measurements, and verified corporate disclosure dates. Across 175 sites with defined exposure, reliance on Sentinel-1B was highly stable between 2020 and 2021 but only modestly correlated with realized coverage loss in 2022. Additional Sentinel-1A observations offset 28.6% of the mechanical loss of B-only observation dates. A separate pilot containing 2,043 radar site-scene measurements supports 319 retrospective test point-quarters. Adding Sentinel-1B features did not improve mean squared production-estimation error in any annual test fold. Financial diagnostics covering 572 disclosures also reveal pre-event trends that prevent a clean causal interpretation. The paper documents the methods, results, and measurement limitations involved in connecting satellite data to economic outcomes. It provides a benchmark for evaluating alternative-data signals without assuming that additional imagery necessarily improves prediction or identifies a market effect. Author: Jonathan Landmann, Stanton College Prep