ESRL Science Making a Difference in Real-time GPS Positioning

GPS water vapor instrument at Cape Canaveral, Florida.

GPS water vapor instrument at Cape Canaveral, Florida.

Scientists at ESRL's Global Systems Division (GSD) are known for transferring their advances in science and technology to operations and have made exciting progress with a product called NOAATrop. NOAATrop was developed to improve GPS positioning, navigation and timing accuracy using real-time weather data. The California Spatial Reference Center (CSRC) at Scripps Institution of Oceanography recently started using NOAATrop for real-time engineering applications in southern California. "This is the first known operational application of a weather model being used to improve high accuracy GPS surveying," said Seth Gutman, NOAA scientist whose years of research at ESRL's Global Systems Division in Boulder have made this advancement possible.

Almost everyone has heard of GPS - the satellite Global Positioning System. GPS was developed by the U.S. Department of Defense to tell us basically three things - where we are, how to get where we want to be, and what time it is right now. This is commonly referred to as "PNT" which stands for positioning, navigation, and timing. GPS is a "dual-use" system in that it has both military and civilian applications. To facilitate the development of civilian applications, the U.S. Government makes the radio signals broadcast by the GPS satellites available to every user free of fees. As a result of this policy, a multi-billion dollar industry has developed to exploit the benefits of GPS for civilian applications. For example, a commercial GPS navigation system in your car tells you where you are on an electronic map that is constantly updated as you move. Computer software in the navigation system tells you the best route to take to get to your destination. Wireless communications between your navigation system and a traffic management system tell you the best route to take to avoid delays or hazards.

According to Gutman, "The largest source of PNT error comes from the slowing and bending of the GPS radio signals as they pass through the Earth's atmosphere" Scientists involved in using GPS for very high accuracy positioning applications, like the ones at NOAA's National Geodetic Survey, developed techniques to treat the atmosphere as a source of noise or measurement error and remove it to improve their GPS positioning accuracy. Gutman continued, "We realized that most of the "noise" actually comes from water vapor in the lower atmosphere, and that improved ways to monitor water vapor would lead to improved weather forecasts and climate monitoring." The operational use of this new and heretofore unavailable water vapor information by the National Centers for Environmental Prediction (Figure 1) has resulted in substantial improvement in U.S. weather forecast accuracy in recent years.

Improvements in precipitable water vapor and humidity predictions by the Operational model due to the assimilation of GPS water vapor measurements.

Fig 1. Assimilation of GPS water vapor measurements into National Weather Service operational weather models significantly improves the accuracy of cloud and precipitation forecasts. These translate into improved severe weather forecasts and warnings, and improved aviation and surface transportation safety.

It did not take long for ESRL scientists to realize that if they could use GPS data to improve weather forecasts, then improved weather model assimilating GPS observations could probably be used to estimate and correct for the atmospheric signal delays that are a major source of GPS PNT error. To verify this, they developed NOAATrop, a model that uses the ESRL Rapid Update Cycle (RUC) weather prediction model to calculate real-time atmospheric correctors for high accuracy GPS positioning applications.

Figure 2 is a map of the Continental U.S. highlighting three regions where the use of the NOAATrop model has been independently evaluated by three universities. The Table in this figure compares GPS positioning accuracy using the NOAATrop model with the accuracy achieved using a conventional correction method known as the Saastamoinen model. This model does not use information about the atmosphere per se, but uses information about how the atmosphere changes on average over North America. The smallest improvement in real-time positioning accuracy (about 9%) is seen in the Southeast U.S. where moisture variability is high and terrain relief is low. The largest improvement (about 25%) is seen in the Western U.S. (California) where moisture variability is low but terrain relief is high. In between is the Great Lakes Region (Michigan) where terrain and moisture variability vary widely. The improvement here is better than 16%.

Improvements in GPS positioning for south east, central and west coast regions.

Fig 2. Improvement in real-time GPS positioning accuracy using the NOAATrop model compared with a conventional correction technique known as the Saastamoinen model (SAAST). The Saastamoinen model is actually a correction based on climatology (expected average conditions) rather than the actual state of the atmosphere that causes the GPS errors and is accurately modeled by NOAATrop.

The RUC forecast model is reinitialized every hour using the latest observations including approximately 300 GPS precipitable water vapor measurements made every half-hour over the United States. "RUC is the highest frequency NOAA weather model and particularly well suited for this unique real-time application," stated Gutman. Figure 3 illustrates how it works.

Initial tests in California using NOAATrop as part of the CSRC Height Modernization project were very positive, yielding 15-25% improvement in accuracy with the greatest improvement shown in height measurements made during rapidly changing weather conditions. "From CSRC's perspective, using NOAATrop increases the accuracy and productivity of field surveys for its user community," said Yehuda Bock, Director of CSRC. Based on their experience, the NOAATrop is currently being used in the Central Valley as a way to increase accuracy and productivity of their road and infrastructure surveys.

Gutman highlighted his lab's cooperation with several agencies on the NOAATrop project. One in particular, the Nationwide Differential GPS Service, has served as part of the nationwide backbone for high accuracy surveys like the ones in California. The exciting thing is that the new development of applications for NOAATrop extends beyond these surveys and leads to cooperation with a number of external partners. "The potential for this product doesn't stop here", said Gutman. National Resources Canada (NRCan) Canadian Geological Survey is working with the California Spatial Reference Center at Scripps on NOAATrop applications to

serve another valuable societal need. By using NOAATrop, scientists who study changes in the land surface due to earthquakes can continuously monitor the positions of GPS receivers in real time with unprecedented accuracy. This has potential application in tsunami warnings along the Pacific Coast on the US and Canada.

Diagram of data flow which improves GPS PNT accuracy.

Fig 3. How weather models can improve GPS PNT accuracy. Water vapor data are estimated at about 300 sites in the U.S. every half hour. Every hour, these measurements are assimilated into an operational weather model called the Rapid Update Cycle (RUC) along with all other measurements made by NOAA observing systems. The RUC model estimates the current state of wind, temperature and moisture at many levels in the atmosphere and predicts the future state of these parameters using the basic laws of physics. NOAATrop takes the model predictions of temperature, pressure and water vapor, computes the GPS errors caused by these parameters, and provides them to surveyors needing very high accuracy. In the example above, U.S. Geological Survey scientists are using GPS to monitor the change in elevation of the surface caused by changes in the magma chamber far beneath an active volcano. Improved survey accuracy results in better volcano eruption warnings and forecasts.

ESRL research efforts have helped attain milestones within the surveying community and the laboratories look forward to broader applications of this technology within and outside of NOAA in the future. They continue to support NOAA's mission goal to serve society's needs for weather and water information as well as the goal to support our nation's commerce with information for safe, efficient, and environmentally sound transportation.