Creation of a Hyperlocal 5-Minute Precipitation Database: A Fusion of Automated Gauges, Manual Observations, and Radar-Estimated Rainfall

The New Jersey Weather and Climate Network is a mesonet of 55 weather observing stations, operated by the Office of the New Jersey State Climatologist. A recent upgrade to this network allows stations to report precipitation and other weather parameters at a temporal resolution of 5 minutes. Following this upgrade, the need and opportunity arose to fuse other streams of daily, hourly, and subhourly precipitation observations with the network. The variety and breadth of multi-sensor precipitation observations throughout the state provide avenues for quality control and improvement of the spatial resolution of the dataset.

A total of 4 NEXRAD sites (KDIX, KDOX, KOKX, KBGM) and 3 TDWR sites (TEWR, TJFK, TPHL) provide radar coverage to monitor precipitation in New Jersey. Storm-total precipitation estimates from these radars have been coupled with in situ rainfall reports from over 30 NWS cooperative network stations, 10 ASOS and AWOS stations, and over 200 active CoCoRaHS observers to supplement the New Jersey Weather and Climate Network. In many cases, precipitation observations from all sensors and gauges for a given point can be interpolated to provide an approximate minute-by-minute reconstruction of a precipitation event. The real value of this method is the fact that calculation of high-resolution storm totals for stations and gridpoints can be automated, providing real-time insight even as the rain is still falling.

The utility of this database has been realized immediately, as the ONJSC has already implemented routines and queries to catch rain gauge clogs and other data quality issues as they occur. Furthermore, the dataset will be used to build tools for high-resolution precipitation analysis and real-time decision-making. Finally, building a long-term, comprehensive database of subhourly precipitation data will eventually allow for easier and more precise applied hydrologic research. This includes more precisely and easily calculating subhourly extreme precipitation estimates and recurrence intervals.