Browse the full Resources Index
Much of the weather data cited in property claims traces back, directly or indirectly, to sources that are freely available to the public. The National Oceanic and Atmospheric Administration and the National Weather Service maintain several archives that anyone with an internet connection can search, and it is worth knowing what they contain before treating a commercial weather report as the only way to access this information. Understanding the public source material also makes it easier to evaluate a commercial report, since most such reports ultimately rest on the same underlying government data rather than on independently gathered observations.
The NCEI Storm Events Database, maintained by NOAA’s National Centers for Environmental Information, compiles storm reports submitted through the National Weather Service network, organized by event type, date, and county. NEXRAD radar archives, also maintained by NOAA, preserve raw radar scans going back decades, which can be retrieved and processed to look at reflectivity and, where dual-polarization data exists, hail-related signatures for a specific time and location. Many states and universities also operate mesonet networks, which are denser grids of automated weather stations than the national ASOS and AWOS network, providing more localized readings in the areas they cover. The Minnesota State Climatology Office, housed within the state’s Department of Natural Resources, is one example of a public entity that compiles and preserves historical weather records for the state, offering another avenue for locating archived data relevant to a specific date and location.
Accessing this raw data is one thing; interpreting it correctly is another. Radar archive files require specialized software to process into usable reflectivity or hail-estimate imagery, and the resulting output still carries the same resolution and inference limitations present in any radar-based analysis. The storm events database depends on what was reported and by whom, meaning gaps in that dataset do not necessarily mean an event did not occur, only that it was not formally reported through that channel. Mesonet coverage, while denser than the national network, is not uniform across every county or state.
Station siting also affects data quality in ways that are not always obvious from a raw dataset. The National Weather Service publishes siting standards addressing how far an anemometer or thermometer should be from buildings, pavement, and other obstructions that can skew a reading, and a station that departs from those standards, whether due to nearby construction or its original placement, can produce systematically biased data for its location. Reviewing metadata about a station’s siting and history, where it is available, is a useful check before treating any single station’s record as fully representative of conditions across a wider area.
For someone without a meteorology or data analysis background, working with raw public data can be more time-consuming and error-prone than it initially appears, which is part of why commercial verification services and forensic meteorologists exist as an intermediate step. Public data remains a legitimate and often useful starting point for understanding what is knowable about a storm event, provided its gaps and its need for careful interpretation are kept in view rather than assumed away.
This article is general education about how the appraisal process commonly works. It is not legal advice, and specific procedures can vary by state and policy.
Russ Lis is a working property insurance appraiser and umpire based in Minnesota, serving clients nationwide. Contact Appraisal Resolution.