← Back to Portfolio

SPATIAL ANALYSIS · PUBLIC SAFETY GIS · WARNING SYSTEM IMPROVEMENT

AWIPS II Location Coverage Analysis

Spatial analysis of the NWS Newport/Morehead City County Warning Area that identified missing census-designated places and introduced volunteer fire department locations as population indicators, increasing WarnGen coverage from 71.9% to 91.4% of the CWA.

---

Problem

The National Weather Service generates warnings that reference specific locations so the public can quickly determine whether they are in an affected area. These location-based warnings depend on a database of recognized places loaded into AWIPS II, the NWS operational forecasting system. If a community is not in the database, warnings issued for that area lack the specific location references that help residents make safety decisions.

At NWS Newport/Morehead City, analysis revealed that only 71.9% of the County Warning Area was within five miles of a recognized location in AWIPS II. That meant over a quarter of the CWA had no location-specific context in generated warnings. The gap was not a minor edge case. It covered populated areas where people lived, worked, and needed actionable information during severe weather.

---

Approach

The analysis started with a straightforward spatial comparison. I buffered every location currently recognized in AWIPS II at four-mile and five-mile distances, then overlaid that coverage against the full CWA boundary. The gaps were immediately visible and substantial.

Next, I identified census-designated places that existed in the Census Bureau’s geographic data but were not loaded into AWIPS II. These were established communities with recognized names that the warning system simply did not reference. Each missing location was classified by its distance from the nearest recognized location: yellow diamonds for places more than four miles away, red diamonds for places more than five miles from any coverage.

The more significant finding came from examining the remaining gaps after adding the missing census-designated places. Even with those additions, areas of the CWA remained uncovered. These were rural areas without incorporated municipalities or census designations, but they were not empty. The presence of volunteer fire departments in these uncovered zones indicated populations significant enough to require organized emergency services but invisible to the warning system.

I proposed adding volunteer fire department locations as recognized places in AWIPS II. The logic was direct: if a community has a volunteer fire department, people live there, and those people need location-specific weather warnings. To my knowledge, this was the first time VFD locations were included in AWIPS II WarnGen at any NWS office in the United States.

---

Deliverable

The final product was a four-tab Esri Story Map presenting the full analysis as an interactive application that NWS staff could explore and share with stakeholders. Each tab built on the previous one to construct the case for the proposed changes.

The MHX Cities Buffer Comparison tab established the baseline by showing the difference between the old AWIPS I city list coverage (blue) and the current AWIPS II coverage (green), making the 28% gap immediately visible. The Current Identified Cities and Cities Missed tab mapped every missing census-designated place against the existing buffer coverage, with color-coded diamonds distinguishing the severity of each gap. The VFDs tab introduced the volunteer fire department concept, showing where VFD locations would extend coverage into areas that even the missing census places could not reach. The Overview of Proposed Changes tab presented the combined result: all proposed additions mapped together, showing coverage increasing from 71.9% to 91.4% of the CWA.

View the interactive story map

---

Outcome

The proposed changes increased WarnGen location coverage from 71.9% to 91.4% of the MHX County Warning Area. Beyond the area coverage improvement, the additions meant that warnings generated for these previously uncovered zones would now include specific location names, giving residents a concrete reference point for determining their proximity to warned areas.

The VFD inclusion represented a novel approach to a problem that existed at NWS offices across the country. The standard process relied exclusively on census-designated places and incorporated municipalities, which systematically underserved rural communities. Using volunteer fire departments as population indicators addressed a structural blind spot in how the NWS identified communities within its warning areas.

---

Why This Matters

This project demonstrates that operational improvements in public safety systems often come from asking practical questions about the data rather than deploying complex technology. The analysis used standard GIS tools: buffers, spatial joins, and area calculations. The innovation was in recognizing that VFD locations served as a proxy for population distribution in areas the census data did not adequately capture.

This was the project that started my path into public safety GIS. The work showed me that spatial analysis applied to operational systems could directly improve public safety outcomes, and that the gap between existing data and operational needs was often larger than the agencies responsible for those systems realized.