When an operator asks how they can verify that a payout was calculated correctly, the answer starts with data source transparency. Every index trigger in a Riskwright policy references a publicly accessible federal observation network. The calculation is reproducible by any party with access to the same data — and that access is free and permanent.
This piece covers the primary networks we use, why we chose them, and how we evaluate data quality before allowing a station to serve as the reference point for a policy.
NOAA ASOS: the backbone for real-time event triggers
The Automated Surface Observing System (ASOS) is a network of approximately 900 weather observation stations operated jointly by the National Weather Service, the Federal Aviation Administration, and the Department of Defense. ASOS stations are located primarily at airports and aviation facilities, which means they are well-maintained, subject to regular calibration, and have near-complete observation records going back to the 1990s in most locations.
ASOS stations are our primary source for wind speed observations, freezing rain accumulation, present weather type, and precipitation. The network reports hourly, with some stations providing one-minute data. For infrastructure triggers — wind speed exceedance during storm events, ice accumulation during freezing rain episodes, 72-hour precipitation totals — ASOS is almost always the designated source.
ASOS data is archived in NOAA's Integrated Surface Database (ISD) and accessible through NOAA's Climate Data Online portal. Any counterparty can pull the same observation record that we used for trigger calculation, with the exact timestamp and QC flag for each reading.
One limitation of ASOS: because stations are at airports, they may not capture conditions well in heavily forested, complex-terrain, or coastal environments. A coastal ASOS station may underestimate inland precipitation by 10–20% in some atmospheric patterns. This is why we assess spatial representativeness before designating any ASOS station as a policy reference.
NOAA GHCN: historical depth for drought and precipitation indices
The Global Historical Climatology Network (GHCN-Daily) is maintained by NOAA's National Centers for Environmental Information and archives daily precipitation and temperature records for approximately 27,000 US surface stations. GHCN includes the longest available instrumental climate records — some US stations extend back to the 1800s, and most have useful records going back 40 to 60 years.
We use GHCN as our primary source for the Standardized Precipitation Index (SPI) calculations. Computing SPI requires a long historical baseline — typically 30 years at minimum — to establish the climatological distribution against which current precipitation is measured. GHCN provides that depth where ASOS, with its shorter record, often cannot.
GHCN quality control is applied at the station level and flags observations that are inconsistent with neighboring stations or with the historical distribution at that location. We review QC flags as part of our station evaluation and exclude stations with high rates of flagged observations from policy use.
PRISM climate grids: spatial interpolation for basis risk assessment
PRISM (Parameter-elevation Regressions on Independent Slopes Model), developed at Oregon State University's PRISM Climate Group, is a gridded climate dataset that interpolates observed station data onto an 800-meter spatial grid. PRISM accounts for topographic effects on precipitation and temperature — the influence of elevation, slope aspect, and coastal proximity on local climate — producing estimates that are often more accurate at specific field locations than the nearest point station reading.
We use PRISM primarily for basis risk assessment rather than as a direct policy trigger source. Before recommending a station, we compare the station's historical record against the PRISM grid cell at the insured's location. If the station and the PRISM cell show consistent behavior over a 20-year period, geographic basis risk is likely low. If they diverge, we investigate why — often a topographic break between the station and the insured's location — and either find a better-positioned station or document the divergence in the policy disclosure.
For drought index calculations in complex terrain, we have developed a blended approach that incorporates PRISM precipitation estimates alongside GHCN station data. This is disclosed in the policy schedule and is reproducible using publicly available tools.
USGS stream gauges: flood trigger reference for infrastructure
For infrastructure operations exposed to riverine flooding — stormwater systems, coastal protection infrastructure, bridge and road networks near waterways — the United States Geological Survey's National Water Information System (NWIS) stream gauge network provides real-time and historical streamflow and water level data at approximately 8,000 active monitoring locations across the US.
Stream gauge-based triggers are particularly useful where the financial impact of a weather event is mediated through river response rather than directly through rainfall. A highway contractor working in a flood-prone basin cares about when the water crests above a certain stage, not just about how much rain fell. A stream gauge trigger set at a specific gage height correlates more directly with the actual operational impact than a precipitation accumulation trigger would.
USGS stream data is accessible in real-time through the NWIS web interface and historically through the NWIS archive. Data quality varies by gauge site; we assess each gauge's record completeness and the history of provisional vs. approved data before specifying it as a policy reference.
The Cooperative Observer Program (COOP): supplementary coverage in rural areas
NOAA's Cooperative Observer Program is a network of approximately 8,500 volunteer weather observers, many of whom have been maintaining daily precipitation and temperature records at fixed sites for decades. COOP stations extend observational coverage into rural and mountainous areas where ASOS and GHCN stations are sparse.
COOP data is less consistently maintained than federal networks — observers are volunteers, and observation schedules and equipment vary — which means quality control is more labor-intensive. We use COOP stations as supplementary sources in specific circumstances: where no federal network station is within 20 miles of an insured's location, or where a COOP station has a documented long record with high completeness. We never use COOP as a primary trigger source without disclosing its quality characteristics in the policy schedule.
Data quality standards and the backup station protocol
Every policy specifies a minimum data completeness threshold for the primary station: if the station reports fewer than 85% of expected observations during a trigger assessment period, the backup station data is substituted. The backup station is identified at policy inception — not after an outage occurs. The interpolation formula for integrating backup station data is documented in the policy schedule.
When an outage occurs or when data quality flags exceed our acceptance threshold, we run the trigger calculation with the backup station data and document which station's data was used in the trigger confirmation notice sent to the insured. The insured can verify independently using the same backup station's publicly archived data.
This transparency has a practical implication: Riskwright can't make up a trigger confirmation or conceal an index reading. The data is public. If there's ever a question about whether a trigger should have fired, both parties can run the calculation independently from the same public archive. That's the point of using federal data sources — it eliminates the information asymmetry that makes traditional insurance disputes so common.
If you want to know which stations would serve as the primary and backup references for a policy at your location, send us your GPS coordinates and we'll run the coverage analysis.
How derived indices are calculated from raw observation data
The raw observation data from ASOS, GHCN, and PRISM isn't used directly as a trigger in most policies. It's processed into derived indices that more precisely capture the exposure being covered. Understanding the distinction between the raw data source and the computed index matters for interpreting your policy schedule correctly.
The Standardized Precipitation Index (SPI) is a common example. GHCN provides daily precipitation observations. To compute SPI-3 for a specific June-August period, those daily observations are summed over the three-month window, then compared against the distribution of all June-August precipitation totals recorded at that station over the 30-year reference period (typically the most recent NOAA 30-year climatological normal). The result is a z-score: SPI-3 of −1.5 means the three-month total was 1.5 standard deviations below the long-run average for that location and period. The formula is documented in the policy schedule so that either party can reproduce the calculation from the raw GHCN data.
Similarly, growing degree day (GDD) accumulation triggers are derived from GHCN or ASOS temperature observations. GDD for a given day is the average of daily maximum and minimum temperature minus a crop-specific base temperature (50°F for corn in most agronomic models). Accumulated GDD from a defined start date signals whether a season is warm enough for expected crop development; frost degree day variants invert the same calculation to measure cold damage exposure.
The role of NOAA CDO and data access architecture
NOAA's Climate Data Online (CDO) portal is the primary access point for GHCN, ASOS, and COOP data in policy-relevant formats. Every trigger calculation in a Riskwright policy is reproducible using CDO-accessible data: the station ID, date range, and variable of interest are specified in the policy schedule, and any counterparty can recreate the exact dataset within minutes. USGS NWIS provides equivalent self-service access for stream gauge data, with real-time readings updated every 15 minutes and historical records freely downloadable.
This data accessibility architecture is not incidental — it is the foundation of why parametric insurance is dispute-resistant. In years of placing infrastructure and energy insurance, the single most common source of claims disputes in traditional policies was disagreement about what the weather conditions actually were. Parametric policies based on federal observation networks eliminate that class of dispute entirely: the data is public, archived, and independently verifiable by both parties.
Limitations of current data infrastructure and where we look for solutions
We're transparent about the gaps in current federal observation coverage. The ASOS network's airport bias creates systematic underrepresentation of inland, forested, and complex-terrain environments. The COOP network fills some of those gaps but at inconsistent quality. PRISM's 800-meter resolution is excellent for regional climate assessment but introduces its own interpolation assumptions in areas with steep elevation gradients.
For operations in those underserved environments, we evaluate whether private weather station networks — particularly those operated by agricultural universities, state climatologist offices, or mesonet programs — can supplement or substitute for federal sources. Several southeastern states maintain mesonet networks with quality standards comparable to GHCN. Where we use a mesonet station, we disclose the operator, quality control procedures, and data access method in the policy schedule, alongside the federal backup station protocol.
We're not saying private weather networks or agricultural IoT sensors are unreliable — many are well-maintained and accurate. What we're saying is that a parametric policy trigger requires an independently verifiable, publicly archived data source, because the transparency and dispute-resistance of the product depends entirely on both parties being able to reproduce the calculation from the same public record. That constraint is the foundation of the transparency claim we make: it holds only if every data source referenced in the policy can be verified by any party with a browser and a station ID.