Summary of Global
Positioning System (GPS) Integrated Precipitable Water (IPW)
Pablo Santos[1]
National Weather Service
Kenneth Carey
Wayne M. MacKenzie, Jr.
Jian Zhang
CIMMS/University of
NOAA National Severe Storms Laboratory
Ralph Ferraro
NOAA/NESDIS
Center for Satellite Applications and Research
James G. Yoe
NOAA/NESDIS
Office of Systems Division
Submitted to: NWA Electronic Journal
April 2007
Decision-makers are increasingly reliant on public and private forecasts for daily planning, hurricane evacuations, fuel distribution, agriculture pricing policy, positioning fire weather assets, and applying road chemicals for winter storms. They can benefit from better forecasts resulting from measurements of atmospheric water vapor from ground-based Global Positioning System (GPS) Integrated Precipitable Water (IPW) retrievals. These GPS-IPW data are being used in weather prediction models and by operational weather forecasters to help produce more accurate analyses and forecasts of the atmospheric moisture patterns over the United States, leading to a wide range of improved forecast applications including severe weather. This paper presents a brief history of the GPS-IPW network along with a summary of recent studies on validations and forecast impacts of these data. Examples of impact in operational forecast scenarios are also presented.
Global Positioning System (GPS) satellite radio signals are slowed as they pass through the Earth’s atmosphere. This delays the arrival time of the transmitted signal from what is expected if there was no atmosphere. The delay in the signal as it travels through the atmosphere originates from both the ionosphere and the neutral atmosphere. The ionospheric-caused delays can be corrected for by using dual-frequency GPS receivers as they are frequency dependent. The delays from the neutral atmosphere, however, are not frequency dependent as they depend on its constituents, which are a mixture of dry gases and water vapor. Using the techniques first described by Bevis et al. (1992, 1994) and Duan et al. (1996), the signal delays caused by water vapor in the troposphere can be estimated and used to retrieve the total column water vapor or integrated precipitable water (IPW). This new technology opened the door for the development of a ground based GPS-IPW network in the 1990s led by the National Oceanic and Atmospheric Administration (NOAA) Earth Systems Research Laboratory (ESRL) Global Systems Division (GSD) (Wolfe and Gutman 2000; Gutman et al. 2004). As of June 2007 the network has grown to nearly 400 sites across the United States, Canada, Mexico, and the Caribbean (Fig. 1).
GPS-IPW complements other systems capable
of measuring atmospheric moisture such as radiosondes, surface-based
radiometers, satellite-based infrared and microwave sensors, research aircraft,
and commercial aircraft [e.g., Aircraft Communication Addressing and Reporting
System (ACARS)]. However, it is not a
substitute as it does not provide information about moisture profiles. Radiosondes provide tropospheric moisture profiles,
but have limited spatial coverage and are only launched twice-daily—in some
countries only once per day.
Surface-based radiometers are capable of high temporal resolution but
are costly, require frequent calibration, and their performance is adversely
affected by the presence of rain.
Satellite-based infrared (IR) and microwave sensors offer planetary
scale coverage, but IR sensors are reliable only in cloud-free regions, and
microwave sensor-based retrievals, although valid in cloudy regions, are most
reliable over oceans (less reliable over land) and have limited temporal
resolution. Aircraft measurements are beginning to provide
moisture observations using the Water Vapor Sounding Systems (WVSS) or Tropospheric Airborne
Meteorological Data Reports (TAMDAR). However, these observations are
limited to commercial operational locations and flight times, and are generally
less continuous than GPS-IPW observations.
In fact, aircraft observations other
than TAMDAR are generally limited to hub airport areas below 15 kft. The GPS-IPW
network provides unattended, continuous, independent, frequent, and accurate
observations of IPW that are unaffected by weather conditions or time of
day. And the cost of each station is
very low; installation cost is usually less than $7,000 if collocated with a
surface meteorological observation station, or around $10,000 otherwise with an
approximate $500 annual operating cost[2]. The main limitations of the GPS-IPW network
are that the IPW retrievals do not provide information about the vertical
distribution of water vapor, and the spatial resolution is limited (although
this is becoming somewhat alleviated by the fast expansion of the network). It also meets essential water vapor monitoring requirements
not met by all other sensors, most significantly its ability to monitor water
vapor under all weather conditions which is critical during potential severe
weather events (United States
Weather Research Program Prospectus Development Team Report; Emanuel et al.
1995). In addition, GPS-IPW accuracy of
1 to 2 mm (Deblonde
et al. 2005) is equal to or
better than integrated radiosonde moisture soundings at a fraction of the cost (Gutman et al. 2005).
But why bother measuring atmospheric
water vapor? Water vapor is one of the most significant constituents of the
atmosphere because it is the means by which moisture and latent heat are
transported in the atmosphere. Water vapor
is also a greenhouse gas that plays a critical role in the global climate
system. This role is not restricted to
absorbing and radiating energy traveling through the atmosphere, but includes
the effect it has on the formation of clouds and aerosols and the chemistry of
the lower atmosphere. Despite its
importance to atmospheric processes over a wide range of spatial and temporal
scales, water vapor is one of the least understood and poorly described
components of the Earth's atmosphere.
Water vapor moves rapidly through the atmosphere, redistributing energy
through evaporation and condensation.
This can occur abruptly over extremely short distances. For this reason, water vapor is
under-observed in time and space, especially during severe weather. This conclusion is supported by multiple
scientific publications, among them a special report on water vapor
in the climate system (1995)[3] published by the American Geophysical
Union (AGU), which states that although the Earth’s “basic operation of the
hydrologic cycle is well known…some details are poorly understood,
mainly because we do not have sufficiently good observations of water vapor.”
The first United States
Weather Research Program Prospectus Development Team Report by Emanuel et
al. (1995) made as one of its key
recommendations “the support of research seeking to determine optimal
combinations of satellite and ground-based remote sensing, aircraft,
balloon, and surface observations as well as the support of key technological
developments such as satellite-borne active sensing techniques, near-field
remote sensing of atmospheric water vapor, and observations from commercial
and, perhaps, pilotless aircraft” as a condition to achieve forecasts
improvements “at the 2-7-day range” which “could have enormous
potential economic benefits but will require greatly improved data over the
oceans and other data sparse areas.” The Global
Climate Observation System (GCOS) workshop report (2006)[4] on the Upper-Air Network includes
recommendations concerning GPS-IPW. The
GPS-IPW network makes it possible to make
observations of IPW with high horizontal resolution (provided the network is
dense enough), high temporal resolution, high accuracy, long-term measurement
stability, and high reliability under all weather conditions. Although at first glance the
applicability of GPS-IPW measurements over oceans is limited, its deployment
across island environments and on platforms such as oil rigs, buoys, and
ships—representative of the oceanic environment in which they are embedded—has
been proposed since they would undoubtedly yield significant benefits (Chadwell
and Bock 2001; Rocken et al. 2005).
The next two sections include brief summaries taken from various published articles on GPS-IPW data validation studies (section 2) and GPS-IPW impacts on forecasts (section 3).
The
errors associated with GPS-IPW estimates are usually determined from
comparisons with other moisture sensing systems, especially radiosondes and
microwave water vapor radiometers (MWR).
NOAA-sponsored studies have been carried out at the Department of Energy
Southern Great Plains (SGP) Atmospheric Radiation Measurement (ARM) Cloud and
Radiation Testbed (CART) Facility near
a. Forecast Models
Assimilation of GPS-IPW data into mesoscale numerical weather prediction (NWP) models has been proven to reduce model 3-h IPW errors by 25% on average over a 3-month period (Smith et al. 2007). This has resulted in increasing improvements in 3-h relative humidity (RH) forecasts below 500 hPa in the south-central United States (Fig. 5) as the GPS-IPW network has increased from 2000-2004. Smith et al. (2007) showed an 8% improvement in 3-h RH forecasts over the entire year, with 10-15% improvement in transition seasons (Fig. 6), and substantial reductions in root-mean–square (RMS) errors of model IPW forecasts across the CONUS region (Fig. 7). Significant improvements in model 3-h Convective Available Potential Energy (CAPE) forecasts, skill scores (ETS) for heavy precipitation events (Benjamin et al. 1998; Deblonde et al. 2005; Smith et al. 2007), and even slight improvements in land-falling hurricane forecast tracks (Fig. 8) have been documented (Macpherson et al. 2007). This overall improvement in NWP performance has resulted in the incorporation of the GPS-IPW data into two models at the National Centers for Environmental Prediction (NCEP), namely, the Rapid Update Cycle (RUC) in June 2005 and the North American Mesoscale (NAM) model in June 2006 (Smith et al. 2007). Fig. 9 illustrates the impact in the operational RUC model of the assimilation of these data starting during the summer of 2005.
These improvements are essential to help NOAA meet its strategic goals[5] of improving severe weather forecasts, aviation forecasts, hydro-meteorological forecasts, and climate forecasts. Additionally, the ability to retrieve atmospheric water vapor content from GPS signals has enabled up to 19% improvement in real time kinematic positioning from GPS signals widely used in surveying techniques (Ahn et al. 2006), with positioning accuracies on the order of centimeters (Bisnath and Dodd 2004). This suggests the development of the GPS-IPW technology has benefited the society at large beyond the weather enterprise. This goes to the core of NOAA’s mission in support of the nation’s commerce.
Additional benefits of the GPS-IPW
network include: 1) quality control of moisture for global radiosonde
observations, which leads to detection of bad soundings and results in improved
moisture observations for NWP, climate statistics, satellite calibration and
validation, and research [Gutman et al. 2005; McMillin et al. 2007; Rama Varma
Raja et al. 2007; see also R. Maddox’s blog (http://www.madweather.blogspot.com)
regarding moisture measurements from the Radiosonde Replacement System (RRS)];
2) verification of satellite and other moisture sensing systems which provides
an independent check on the quality of remotely sensed measurements from
satellites and/or in situ measurements from radiosondes (Birkenheuer and Gutman
2005); and 3) improved
situational awareness to forecasters leading to better short-term regional
warnings and forecast services that could save life and property [personal
communications with Science and Operations Officers at National Weather Service
(NWS) field offices].
As suggested in the previous section, the
use of the GPS-IPW data has resulted in both better performances of operational
NWP models during severe weather events and better situational awareness
leading to such events in the field. The
next three examples illustrate this.
Fig. 10a depicts the
severe weather reports associated with an outbreak across northern
On 15 May 2006, severe thunderstorms
developed across South Florida resulting in numerous reports of penny-sized to
golf-ball-sized hail covering roadways and occasionally breaking through wind
shields in cars. Reports of wind gusts
in excess of 60 mph were also common (Fig. 10b). Excerpts from the Area Forecast Discussion
issued at 951 AM EDT by the NWS in
“.UPDATE…CONVECTIVE PARAMETERS CALCULATED
WITH MORNING SOUNDING DATA LOOKING VERY IMPRESSIVE. STEEP MID-LEVEL LAPSE RATE WITH AN AFTERNOON
LIFTED INDEX OF -11C...CAPE OVER 5000 J/KG...ELEVATED DRY LAYER...*RAPIDLY INCREASING LOW-LEVEL MOISTURE*...GOOD SURFACE HEATING AND SEABREEZE DEVELOPMENT
JUST SOME OF THE MORE PROMINENT FEATURES
THAT WILL SET US UP FOR SOME STRONG STORMS THIS AFTERNOON. WILL UPDATE HAZARDOUS WEATHER OUTLOOK TO RAMP
UP SEVERITY POTENTIAL JUST A TAD.”
Although not specifically mentioned, the
comment in bold was also based on the time series plot of GPS-IPW shown in Fig. 12. IPW increased
through the morning hours from around 1 inch around 1200 UTC in the
The
third example illustrates the impact of assimilating the GPS-IPW data on the Japan
Meteorological Agency mesoscale model precipitation forecast for a heavy
precipitation event that occurred on 27 Aug 1998 in the main
Drawing upon previous work, this paper summarizes how the GPS-IPW data have become very important for more accurate analysis of the moisture field in the atmosphere. Its accuracy compares exceedingly well to other more conventional platforms, and is even used as a tool for quality controlling/cross-calibrating radiosonde IPW data. It is well documented that the data have a substantial positive impact on NWP models resulting in better forecasts of severe weather events and heightened situational awareness in the field, leading to better forecast and warning services. These data, with their high temporal and increasing spatial resolution, complement rather than supplant radiosondes and other devices capable of measuring moisture profiles. The co-existence of this wide array of sensors and associated networks clearly supports the recommendation made by the first United States Weather Research Program Prospectus Development Team over a decade ago, namely, to support research seeking the optimal combination of space and ground based sensors for better monitoring of moisture fields and their evolution.
Ahn, Y. W., G. Lachapelle, S. Skone, S. Gutman, and S. Sahm, 2006: Analysis of GPS RTK performance using external NOAA tropospheric corrections integrated with a multiple reference station approach, GPS Solutions, 9, 1–16.
Basili, P., S. Bonafoni, V. Mattioli, P. Ciotti, and
N. Pierdicca, 2004: Mapping the
atmospheric water vapor by integrating microwave radiometer and GPS
measurements. IEEE Trans. Geosci. Remote
Sens., 42, 1657–1665.
Benjamin, S., T. Smith, B. Schwartz, S. Gutman, and D. Kim, 1998: Precipitation forecast sensitivity to GPS precipitable water observations combined with GOES using RUC-2. Preprints, 2th Conf. on Numerical Weather Prediction, Amer. Meteor. Soc., Phoenix, AZ, 249–252.
Bevis, M., S. Businger, T. A. Herring, C. Rocken, R. A. Anthes, and R. H. Ware, 1992: GPS meteorology: remote sensing of the atmospheric water vapor using the global positioning system. J. Geophys. Res., 97, 75–94.
Bevis, M., S. Businger, S. Chiswell, T. A. Herring. R. A. Anthes, C. Rocken, and R. H. Ware 1994: GPS meteorology-mapping zenith wet delays onto precipitable water. J. Appl. Meteor., 33, 379–386.
Birkenheuer, D. and S. Gutman, 2005: A comparison of GOES moisture-derived product
and GPS-IPW data during IHOP 2002. J.
Bisnath, S. and D. Dodd, 2004: Analysis of the utility of NOAA-generated tropospheric refraction corrections for the next generation nationwide DGPS service. Proceedings of ION GNSS 2004, Long Beach, CA, 17th International Technical Meeting of the Satellite Division of The Institute of Navigation, 1288–1297.
Chadwell, C. D. and Y. Bock, 2001: Direct estimation of absolute precipitable water in oceanic regions by GPS tracking of a coastal buoy. Geo. Res. Let., 22, 3701–3704.
Deblonde G., S. Macpherson, Y. Mireault, and P. Héroux, 2005: Evaluation of GPS precipitable water over Canada and the IGS network. J. Appl. Meteor., 44, 153–166.
Duan, J. M., and coauthors, 1996: Remote sensing atmospheric water vapor using the Global Positioning System. J. Appl. Meteor., 35, 830–838.
Emanuel, K., and coauthors, 1995: United States Weather Research Program Prospectus Development Team Report. Online report located at: http://www.esrl.noaa.gov/research/uswrp/PDT/PDT1.html.
Emardson, T. R., J. Johansson, and G. Elgered, 2000: The systematic behavior of water vapor estimates using four year of GPS observations. IEEE Trans. Geosci. Rem. Sens., 38, 324–329.
Guerova, G., E. Brockmann, J. Quiby, F. Schubiger, C. Matzler, 2003: Validation of NWP mesoscale models with Swiss GPS network AGNES. J. Appl. Meteor., 42, 141–150.
Gutman, S.I., S.R. Sahm, S.G. Benjamin, B.E. Schwartz, K.L. Holub, J.Q. Stewart, and T.L. Smith, 2004: Rapid retrieval and assimilation of ground based GPS precipitable water observations at the NOAA Forecast Systems Laboratory: Impact on weather forecasts. J. Meteor. Soc. Jap., 82, 351–360.
Gutman, S.I., Facundo, J. and Helms, D., 2005: Quality control of radiosonde moisture observations. Preprints, Ninth Symposium on Integrated Observing and Assimilation Systems for Atmosphere, Oceans, and Land Surface (IOAS-AOLS), San Diego, CA, Amer. Meteor. Soc., CD-ROM, 11.7
Haas, J., H. Vedel, M. Ge, and E. Calais, 2001: Radiosonde and GPS zenith tropospheric delay (ZTD) variability in the Mediterranean. Phys. Chem. Earth (A), 26, 6–8.
Macpherson, S., G. Deblonde, J. Aparicio, and B.Casati, 2007: Impact of ground-based GPS observations on the Canadian Regional Analysis and Forecast System. Preprints, 11th Symposium on Integrated Observing and Assimilation Systems for the Atmosphere, Oceans, and Land Surface (IOAS-AOLS), San Antonio, TX, Amer. Meteor. Soc., CD-ROM, 5.5.
McMillin, L. M., J. Zhao, M.
K. Rama Varma Raja, S. I. Gutman, and J. G. Yoe, 2007: Radiosonde humidity corrections and potential
atmospheric infrared sounder moisture accuracy. J. Geophys. Res., 112,
D13S90, doi:10.1029/2005JD006109.
Nakamura, H., K. Koizumi,
and N. Mannoji, 2004: Data assimilation
of GPS precipitable water vapor into the JMA mesoscale numerical weather
prediction model and its impact on rainfall forecasts. J. Meteor. Soc. Jap.,
82, 441–452.
Rama Varma Raja, M. K., S.
I. Gutman, J. G. Yoe, L. M. McMillin and J. Zhao, 2007: The validation of AIRS retrievals of
integrated precipitable water vapor using measurements from a network of ground
based GPS receivers over the Contiguous United States, J. Atmos. Oceanic
Technol., in press.
Revercomb, H. E., and coauthors, 2003: The ARM program's water vapor intensive
observation periods. Bull. Amer. Meteor. Soc., 84, 217–236.
Rocken, C., J. Johnson, T. V. Hove, and T. Iwabuchi, 2005: Atmospheric water vapor and geoid measurements in the open ocean with GPS. Geo. Res. Let., 32, 1-3.
Smith, T. L., S. G. Benjamin, S. I. Gutman, and S. R. Sahm, 2007: Short-range forecast impact from assimilation of GPS-IPW observations into the Rapid Update Cycle. Mon. Wea. Rev., 135, 2914-2930.
Tregoning, P., R. Boers, D. O’Brien, and M. Hendy, 1998: Accuracy of absolute precipitable water vapor estimates from GPS observations. J. Geophys. Res., 103, 28701–28710.
Westwater, E. R., Y. Han, S. I. Gutman, and D. E. Wolfe, 1998: Remote sensing of total precipitable water vapor by microwave radiometers and GPS during the 1997 water vapor intensive operation period. Proceedings of IGARSS’98, Seattle, WA, IEEE Geoscience and Remote Sensing Society and Cosponsors, 2158-2162.
Wolfe, D. E. and S. I. Gutman, 2000: Development of the NOAA/ERL Ground-Based GPS
Water Vapor Demonstration Network: Design and Initial Results,
The authors wish to thank Seth Gutman,
from NOAA’s ESRL GSD for providing valuable background information during the
preparation of this manuscript, as well as members of the NWA Council for their
valuable input. We would also like to
acknowledge Stan Benjamin, an anonymous reviewer, and Matthew Bunkers for their
invaluable comments.
Figure 1.
NOAA GSD GPS-IPW network as of August 2007 (http://gpsmet.noaa.gov).
Figure 2.
Comparison of 3600 GPS-IPW retrievals and radiosonde IPW over 3 years at
the ARM CART facilities near
Figure 3. Comparison of collocated radiosonde and GPS-IPW measurements during IHOP-2002. From Birkenheuer and Gutman (2005).
Figure 4. Comparison of GPS-IPW to microwave water vapor radiometer (MWR; top) and radiosonde (bottom) measurements taken from Tregoning et al. (1998) for different networks across Australia. Standard deviations between GPS and MWR ranged from 1.3 to 2.4 mm and between GPS and radiosondes standard deviations ranged from 1.5 to 2.7 mm; the average was ~ 2 mm in both cases.
Figure 5.
Normalized forecast impact for the 3-h relative humidity (RH) forecast
error (using RUC60) from assimilation of GPS-IPW data (from Smith et al.
2007). Impacts at 850, 700, 500, and 400
hPa averaged by year for 1999-2004 are shown.
Forecast error is assessed by computing forecast minus observed RH
difference with radiosonde observations at 17 stations in the south-central
United States. Normalized forecast
impact is proportional to the ratio of the difference between the
root-mean-square (RMS) error (mm) of the forecast with no GPS and the forecast
with GPS to the difference between the forecast with no GPS and the verifying
analysis.
Figure 6. As in Fig. 5 except by month for the years 2000–2004 using (a) 850 hPa and (b) 700 hPa (from Smith et al. 2007).
Figure 7.
RMS error (mm) for RUC20 IPW forecast grids against GPS-IPW observations
using 275 GPS sites across the CONUS for the March to May 2004 period (from
Smith et al. 2007).
Figure 8. Two-day storm tracks for Hurricane Frances [from Macpherson et al. (2007)]. The 6-h positions are plotted from 1200 UTC 5 Sep 2004 to 1200 UTC 7 Sep 2004 (black asterisks). Red (blue) asterisks represent the experiment with (without) GPS-IPW.
Figure 9.
Impact on operational RUC IPW forecasts across the CONUS since ingesting
GPS-IPW data during the summer of 2005 when they started to assimilate the data
into the model (NOAA ESRL GSD 2005 GPS-Met Technical Review online report
at: http://gpsmet.noaa.gov).
(a)
(b)
Figure 10. Severe weather reports for (a) 20 Apr 2004 and (b) 15 May 2006.
Figure 11.
(a) 3-h forecast CAPE valid 0000 UTC 21 Apr 2004 from the 20-km RUC
without GPS-IPW and (b) with GPS-IPW.
Intervals in color legend are 250 J kg-1 (from Smith et al.
2007).
Figure 12. Time series plot of GPS-IPW across selected sites in South Florida covering the period from 12 May to 19 May 2006. Notice the rapid increase in moisture across the area beginning early on 15 May.
Figure 13. Observed RAOB (top) at Miami, FL valid at 1200 UTC 15 May 2006 and RUC13 diagnostic sounding (bottom) valid at 1500 UTC 15 May 2006 (from http://rucsoundings.noaa.gov). Notice moistening and warming (wind veering with height) at low levels with lapse rates becoming steeper during the period. In fact from the 1200 UTC RAOB the 850mb - 500mb temperature index was 27.7C and from the diagnostic sounding valid at 1500 UTC the same index was 29C. Also, mid level temperatures went down from -11C to around -12 through the period. As it is, these numbers represent 5 to 6 degrees below normal for 500mb temperatures across South Florida for this time of the year.
Figure 14. Radar observed 3 hours precipitation on the main island of Japan on 27 Aug 1998 (top); Japan Meteorological Agency (JMA) mesoscale model 3 hours precipitation forecast without the assimilation of the GPS-IPW data for the same period (middle); and the model’s precipitation forecast with the GPS-IPW data assimilated (from Nakamura et al. 2004).
[1] The views expressed herein are those of the author and do not necessarily reflect the position of the National Weather Service.
[2] Per verbal communication with Seth Gutman from NOAA ESRL GSD.
[3] Special report on water vapor in the climate system located at: http://www.agu.org/sci_soc/mockler.html.
[4] The Global Climate Observation System (GCOS) workshop report on the Upper-Air Network available at: http://www.oco.noaa.gov (2006).
[5] NOAA Strategic Planning Office Website located at: http://www.ppi.noaa.gov/spo.htm.