Densifying the in-situ surface observing network: The Citizen Weather Observer Program, its role in observation and verification, and avenues for enhancement

The United States National Oceanographic and Atmospheric (NOAA)'s National Weather Service (NWS) traditionally derives the majority of its surface observing data from over 1900 airport-based Automated Surface Observing Systems (ASOS) installations, operated and maintained by a variety of organizations, including the Federal government, various State entities, non-governmental organizations, and private entities. Augmentation of this network, by densifying it by a factor of at least 2, could provide an invaluable tool to the NWS by providing its forecasters with key in-situ information as well as weather event verification data.

The Citizen Weather Observer Program (CWOP) comprises some 5700 total registered members, with some 4200 active. Although the distribution of the group is worldwide, the vast majority are within the Continental United States. A recent query of active CWOP sites yielded over 1900 CONUS reporting stations.

Clearly, doubling the number of stations routinely reporting surface observations is of some value to the Weather Enterprise, if properly managed. CWOP is operated by a group of volunteer administrators who oversee registration, site metadata collection, and observation quality checking. Data are routed via the internet to a volunteer network of servers, and once every 15 minutes, uploaded to NOAA/Forecast Systems Laboratory systems when they are integrated into the Meteorological Assimilation Data Ingest System (MADIS). Within the MADIS infrastructure, the data are again quality checked, and then made available in near-real-time, with the quality parameters incorporated in the NetCDF (common data format) distribution files.

CWOP data already enjoy a favorable history of operational usage. NWS personnel have used CWOP data to augment ASOS in forecasting freezing conditions in Florida, forecasters in Nevada augmenting sparse observation networks, and to augment fire-weather operations. Clearly, network densification has its proponents.

Use of volunteer data, however, has potential drawbacks. Relatively strict criteria exist for most NWS official observing systems, and volunteers are often unaware of such requirements or unable to meet them. Collection of metadata associated with the location and siting of volunteer systems has proven to be a time-consuming task, when one seeks to verify site location and installation specifics. Further, some information is inadvertently provided in error, and correction can prove difficult. Finally, most weather instrumentation requires periodic maintenance and calibration, and volunteers are less likely than formal operational agencies to carry out potentially costly and relatively frequent maintenance operations unless a problem can be identified and its error quantified.

CWOP has a process for improving siting, based on a document (Chadwick, Helms, at al) discussing good practice and providing examples of acceptable and unacceptable implementations. This document was prepared by several of the administrators, and offered for review to the academic and operational community; few errors were found and all were discussed and corrected.

Quality checking, outside of the realm of the FSL MADIS QC, has been undertaken by one of the authors (Gladstone). Specific CWOP participating sites with potential data anomolies are notified, and their data and the associated issues are published on a website with analysis of a first-approximation of the potential problem. This provides a starting point for the volunteer observers to understand the issues of calibration and measurement, an engineering endeavor often poorly understood even by those active in the field.

Future enhancements of the program should focus on automation of the process of registration, site and instrument metadata entry, metadata validation, anomoly registration and reporting, and an improved delivery schedule to NOAA/FSL for systematic QC and dissemination. This set of efforts will build on the efforts already underway, while reducing the current human intervention requirements, and improving the speed of registration and feedback for potential errors.

This presentation will detail specifics of the utility of CWOP in operational instances, and make a case for a more formal recognition of the program by NWS. It will also detail the process of automating and centralizing the tools, functions and data repositories for CWOP operations, with a long-term view of enhancing both NWS and FSL (and other organizations') uses of CWOP data. The potential exists to extend the system to be outlined here into other cooperative systems of observing networks with little effort.