Lower-Cost GPS Met Station Design for use in Dense Network Slant Path GPS-Met Estimates of Tropospheric Wet Delay and Precipitable Water Vapor

GPS-Meteorology is a technique for determining the water vapor content of the atmosphere by analyzing signal propagation delays occurring between a set of signals transmitted by a constellation of GPS satellites and the corresponding set of signals received on the ground and collected by a GPS receiver antenna. It has been shown that this method produces spatially-averaged estimates of the atmospheric water vapor content to within 2% accuracy (S. Hagemann, et al. 2002). As the tropospheric delay consists of two components, hydrostatic delay or dry delay and wet delay, water vapor varies mainly as a function of wet delay. A single GPS-Met station averages the wet delay over a column above the location of the station called the Zenith Wet Delay. This quantity, in turn, is converted into an estimate of the zenith path-integrated precipitable water vapor contained within the column above the GPS-Met station's antenna. While this spatially-averaged water vapor quantity is useful for storm forecasting and other applications, there remains the desire to achieve higher temporal and higher spatial resolution observations of water vapor as well, beyond that which can be achieved using a single GPS Met station. Slant-path GPS Meteorology has been proposed as a technique to achieve better spatial resolution of the water vapor field. This approach relies on a densely-placed mesh network of GPS-Met stations and a technique to simultaneously process the signals received from all the stations so as to trace the precise ray paths of the signals from each of the satellite to each of the the ground-based receivers. The individual ray paths at each epoch or slant total delay(STD) can be used to estimate the 3D structure of the water vapor field. In Germany a mesh of existing GPS Met stations has been tested to measure slant delay and compare the results with Zenith Wet Delay (ZWD) measurements (M. Shangguan, et. al, 2013). The GPS stations in this network were spaced, on average, 40 km apart. While promising, these results did exhibit significant variation when the spatially-resolved water vapor results were compared with radio-sonde data. The authors of that study hypothesize that a denser GPS Met network having receiver spacing closer than 20 km may be necessary to create more accurate data using this technique. This increase in density has the potential to create cost challenges for those interested in deploying such networks. This paper describes an approach to reduce the cost of these sensors. As a starting point in sizing the number of stations needed in a slant-path GPS network, we assume that a network of 9 nodes, spaced 20 km apart, defines the minimum installation that would support testing of the slant-path GPS concept. The GPS-Met stations typically used today for zenith estimates of water vapor tend to cost ~$30K- $40K, plus installation costs. Therefore, the installation of a denser mesh comprised of 9 stations requires a capital cost of at least $270K to $360K. We have set ourselves the goal of assembling a GPS Met station using a set of equipment that would have a total cost of $10K - $15K. Therefore, a 9-node installation would require $90k - $135k, which is a considerable savings. System Design Approach: We achieved this lower cost design by doing the following: The system was designed to meet all GPS-Met station requirements defined by WMO(World Meteorological Organization) and at the same time keeping the entire system cost as low as possible using non-standard GPS-Met station hardware integrated with custom and open source software. Test results: The Integrated Precipitable water vapor obtained for 6 days during Summer 2014 is shown in figure 1. The corresponding LC phase residuals RMS error had a range of 12.1 to 17.1. The LC phase residual RMS error is a good measure of the reliability of the GPS station (Andrzej et. al. 2013). As can be seen in the graph there is a relatively high value of water vapor during days 184-186(UTC) of 40-50 mm. This was the period when Hurricane Arthur hit Massachusetts during the 4th of July weekend. Initial results from our prototype lower-cost GPS-Met station exhibit an average RMS error of 14.1 when compared to an error rate of 5.2 to 11.1 from current industry standard equipment. We plan to install a set of these sensors into the Dallas/Fort Worth/Arlington Texas metroplex area to begin experimenting with the Slant path concept beginning Fall 2014. Acknowledgement: Funding for this work was provided by the Jerome M Paros Fund for Measurement Science.