| GPS 'Noise" Benefits Weather Forecasting
Signal delays measure atmospheric water vapor
By Michael W. Michelsen, Jr. Anyone with an interest in the weather should be aware that an innovative application of the global positioning system provides the National Oceanic and Atmospheric Administration (NOAA) with a new perspective on a critical weather parameter-water vapor.
"In May 1995, a report produced by the U.S. Weather Research Program identified a national need for a reliable, low-cost system for measuring atmospheric water vapor," explained M.J. Post, Ph.D., chief of the Systems Demonstration and Integration Division of NOAA's Environmental Technology Laboratory (ETL) in Boulder, Colorado.
According to Post, water vapor is one of the most important factors in weather because it releases tremendous amounts of energy as it condenses to form clouds and precipitation--often producing severe weather. Water vapor is also the most important greenhouse gas and plays a dominant role in the global climactic system. As a result, measuring atmospheric water vapor is essential for accurate weather forecasting, climate monitoring, and research. But the lack of accurate and timely information adversely affects our ability to make improvements in these critical areas today.
There has always been great interest in mitigating the adverse impact that atmospheric water vapor has on GPS surveying, but it wasn't until 1992 that the first use of GPS as a meteorological remote sensing tool was proposed.
Since then, many researchers have advocated the use of ground- based data from GPS satellites to measure water vapor in the atmosphere. This is truly an example of the adage that one person's "noise" is another person's "signal." Water Vapor Measurement: Then and Now
Most of the water vapor that surrounds the earth is in the troposphere, that portion of the atmosphere which extends outward 7 to 10 miles from the earth's surface. In the U.S., measurements of tropospheric water vapor are routinely made by the National Weather Service using instruments carried by weather balloons launched from more than 70 sites twice daily.
"These measurements have a lot of utility, but also some problems," said Russell Chadwick, Ph.D., chief of the Demonstration Division of FSL that operates the first network of GPS receivers dedicated to atmospheric observations.
"Weather balloons measure the vertical profile of water vapor, which can be integrated to determine the total amount--a quantity measured directly by ground-based GPS receivers. But the balloon sensors are not always very accurate, are expensive to use, and do not provide the frequent measurements needed for improved weather forecasting. The meteorological community needed to find something that reduces costs, decreases time between measurements, and improves accuracy. That's why we started looking at GPS-based systems. We had a lot of evidence that a ground-based system based on GPS signals would work."
The major advantages of using the GPS-based water vapor remote sensing system "are that GPS works continuously under all weather conditions, does not require calibration, and is inexpensive compared to other water vapor monitoring systems such as radiosondes, microwave radiometers, and satellite sensors," said Daniel Wolfe, a meteorologist at NOAA/ETL. How Are GPS Receivers Used to Measure Integrated Water Vapor?
According to Seth Gutman, a systems analyst in FSL's Demonstration Division and the GPS integrated precipitable water vapor (GPS-IPW) project leader, GPS satellite radio signals are slowed by the Earth's atmosphere, which results in a delay in the arrival of a signal as compared to propagation in free space. It is possible to correct for the frequency-dependent portion of the atmospheric delay due to the ionosphere by using dual-frequency GPS receivers, as is routinely done today with high accuracy. The delay due to the neutral (non-ionized) portion of the atmosphere depends on the amount of dry gases and water vapor. The signal delays introduced by these two parts of the neutral atmosphere are called the hydrostatic delay and wet display, respectively. In practice, the signal delays from each satellite in view caused by the neutral atmosphere are mapped to the vertical using a mapping function and combined to give the zenith tropospheric delay (ZTD=ZHD+ZWD). At sea level, ZTD has a magnitude of amount 250 cm, to which the hydrostatic and wet components contribute about 97% and 3%: the approximate mass percentages of the total dry air and water vapor.
By making a precise measurement of atmospheric pressure at each site, scientists can easily calculate ZHD. Subtracting this from observed tropospheric delay yields the signal delay caused by the water vapor in the atmosphere. The integrated precipitable water vapor (IPW), or depth of liquid water that would result if all water vapor in a vertical column of the atmosphere is condensed, is then calculated from the ZHD. A record of average surface pressure, surface temperature, ZTD, and IPW is available every 30 minutes. Data at Hand
NOAA's active interest in GPS-IPW started with two experiments: GPS- STORM, conducted in the spring of 1993, and GPS-WISP, conducted in the winter of 1994. These experiments provided information that led to the installation of the first three NOAA GPS-IPW sites including the National Weather Service forecast office in Denver, and the NOAA Profiler Network (NPN) sites at Platteville, Colorado, and Lamont, Oklahoma. As of August 1997, there were 17 operating GPS-IPW systems. Most of these are located at NPN sites in the Central U.S., although the two newest are installed at sites near Talkeetna and Glennallen, Alaska.
At a typical NPN site, a Trimble GPS antenna (either a 4000ST Ground Plane antenna or an L1/L2 Compact Dome) is mounted about 2.5m above the ground at the corner of a fence surrounding the radar wind profiler antenna. Long-term monitoring of these antenna positions has shown this installation to be extremely stable. The GPS receivers (both Trimble 4000 SSE or 4000 SSi are used) are located in a nearby electronics shelter that provides power and communications to the receiver.
The raw L1 and L2 carrier phase observations from the GPS receivers, along with temperature and pressure measurements made by surface meteorological sensors, are transmitted every 30 minutes to a computer in Boulder, Colorado, using the existing NPN data communication system.
GPS data are processed as soon as satellite orbits of sufficient accuracy (better than 25 centimeters RMS) are available. Currently, this takes about 24 hours, but efforts are underway to reduce this to real-time, and much has already been accomplished toward this end. "The timeliness of data is very important for local and regional weather prediction," said Post. "Obviously, the more often we get current and reliable data, the better we can detect changes in the atmosphere and improve weather forecasts." The GPS Hub
As previously mentioned, data collected by the Trimble GPS receivers at NPN sites are transferred automatically to a computer located at the Profiler Control Center (PCC) in Boulder, called the GPS Hub. The GPS Hub performs the following functions:
¥ Remotely monitors the status of all GPS receivers
¥ Automatically downloads GPS data every 30 minutes
¥ Automatically receives the necessary surface meteorological data acquired at each site during the data acquisition period for the NOAA Profiler Network Hub.
¥ Takes corrective action in the event of malfunction
¥ Notifies an operator in the event of an unrecoverable error
¥ Stores the raw GPS data and converts it to the Receiver Independent Exchange Format (RINEX)
¥ Archives the RINEX data
¥ Receives and stores processed water vapor data
In addition, the GPS Hub tracks system performance and engineering factors such as reliability and maintainability. Data availability is impressive: on average, 98% of all possible GPS data are captured and processed by the Hub.
To date, most major system failures have been caused by lightning strikes. When this occurs, a receiver can usually be replaced within four days, and the site returned to full operation with less than an hour's labor. The Future of GPS in Weather Prediction
In 1995, in cooperation with the National Geodetic Survey, the FSL proposed a program to the U.S. Coast Guard that would allow FSL to install surface sensors at all differential GPS sites operated by the USCG and the U.S. Army Corps of Engineers. FSL could then use existing USGS receivers, as well as the communications infrastructure for downloading and reporting of data, to expand the water vapor network at only a small additional cost.
A prototype meteorological sensor package necessary for properly interpreting the GPS delays was developed by FSL and is called a GPS Surface Observing System (GSOS). This package was completed in January 1996; it costs approximately one-third the cost of a GPS receiver and can be deployed in conjunction with any GPS receiver in the field, to send all necessary data to the FSL hub.
"Analysis of the variability in GPS total delay to retrieve atmospheric water vapor has been proven to be an excellent technique for continuously measuring water vapor," Post commented. "This project has been so successful that we hope NOAA will soon deploy more than 200 GPS water vapor systems in the national network across the United States. Perhaps then meteorologists and modelers will use this exciting new remote sensing technique to markedly improve both short-term and longer-term forecasts, at low cost." About the Author:
Michael W. Michelsen, Jr., is a marketing writer for Trimble Navigation Ltd., Sunnyvale, California. He may be reached at 800-827-8000 ext. 3658. Back |
