GIS Applications in Meteorology, or Adventures in a ... - AMS Journals

10 downloads 106692 Views 291KB Size Report
171. FEBRUARY 2005. AMERICAN METEOROLOGICAL SOCIETY ... The NWS Office of Science and Technology/Systems ... to speed in GIS education. No job ...
GIS Applications in Meteorology, or Adventures in a Parallel Universe BY

SCOTT T. SHIPLEY

I

n my report to the Advanced Weather Interactive hydrometeorology. Instead of the variety of unique Processing System (AWIPS) program in 1994, I data formats extant in the meteorological community said, “GIS is too slow and will never have value for today, these data are revisited as common geographic meteorological data processing.” Despite that sage ob- objects—either as points, lines (arcs), polygons, or servation, GIS strategies have become increasingly popular in meteorological applications, starting with the adoption of open GIS data formats by AWIPS, and recently including the development of GIS applications at NWS River Forecast Centers and Weather Forecast Offices. GIS strategies are now being extended into the imaging stronghold of satellite meteorology and remote sensing. Key to the success of a GIS strategy is the adoption of standards for data interchange and interoperability, especially for Web applications (e.g., Open GIS Consortium, www.opengis.org). A FIG. 1. Rawinsonde tracks reveal extent of spatial offsets at higher altitudes. Here, 100-hPa reanalysis by GIS (raster background) demonstrates second factor is community accepa 5-m s−1 sampling error if these observations are assumed to be located tance and adoption of common over the release point (squares). Intrinsic GIS functions were used, and tools across disciplines (e.g., Com- no code was written. mercial Joint Mapping Toolkit or C/JMTK, www.cjmtk.com). The GIS community is now evolving very rapidly with rasters as images or grids. Intrinsic GIS and externally interfaces to Java, and, just recently, ESRI released a defined user functions are used to analyze these data LINUX version of ArcEngine. objects, demonstrating that GIS is more than just a Features of commercial, off-the-shelf GIS can be “tool for drawing maps”—the most common misconexploited to address traditional problems and tasks in ception. Our hydromet community has evolved a “doit-yourself” strategy to develop the tools it has needed AFFILIATIONS: SHIPLEY—Department of Geography, George to display and process hydromet information, primaMason University, Fairfax, Virginia, and Raytheon, Upper rily because the hydromet data-processing tools did Marlboro, Maryland not exist or were cost prohibitive to acquire. We now CORRESPONDING AUTHOR: Scott Shipley, Senior Scientist, face a decision either to embrace GIS as is, or adopt Raytheon, 1616 McCormick Dr., Upper Marlboro, MD 20774 its strategies in our emerging software systems—the E-mail: [email protected] main difference in outcomes being program risk and DOI: 10.1175/BAMS-86-2-171 cost, especially the logistical cost of software mainte©2005 American Meteorological Society nance and documentation. AMERICAN METEOROLOGICAL SOCIETY

FEBRUARY 2005

| 171

A few examples illustrate where a commercial GIS could meet or exceed the content of discipline/organization-specific applications, commonly referred to as “stovepipes.” Consider the GIS analysis of rawinsondes released at 1200 UTC on 29 December 1998, shown in Fig. 1 on the previous page, clearly revealing a polar jet shearing the raob tracks at higher altitudes. Intrinsic (“out of the box”) GIS functions were used to perform a reanalysis of the wind speed at 100 hPa, using the common assumption that the rawinsondes are located over their release points. This analysis yields a sampling error ~5 m s−1 in wind speed for regions of significant vertical shear, and ~20° in wind direction at the entrance and exit regions of the jets. This sampling error is similar in magnitude to

the measurement error assumed for rawinsondes. Tests underway indicate that GPS sondes should vindicate rawinsonde accuracy. The NWS Office of Science and Technology/Systems Engineering Center has adapted a NEXRAD-toShapefile conversion utility to support NWS radar siting analyses using ESRI ArcView GIS with the Spatial Analyst and 3D Analyst extensions. As shown in Fig. 2, the resulting ArcView GIS project computes and visualizes beam propagation and occultation over terrain databases and supports the correlation of radar coverage with geographic census datasets. This GIS application is used to identify the best locations for potential radar installations and has been modified to depict Federal Aviation Administration weather radars. The con-

USING ARCGIS TO STUDY THE CORRELATION BETWEEN LIGHTNING STRIKE DENSITY AND TERRAIN ELEVATION ALEX J. DECARIA AND MICHAEL J. BABIJ Department of Earth Sciences, Millersville University, Millersville, Pennsylvania BY

With the widely used, off-the-shelf software package, ArcGIS, we investigated the correlation between cloudto-ground (CG) lightning strike density and terrain elevation in southeastern Pennsylvania. Lightning strike data for the region from 1995 through 2001

were obtained at no charge, under their research policy, from Vaisala, Inc., owners of the National Lightning Detection NetworkTM (NLDN). The NLDN is an array of over 100 sensors spread throughout the continental United States that uses time-of-arrival

FIG. SB1. Average annual lightning strike density (km−2 yr−1) for southeastern Pennsylvania from 1995 to 2001.

172 |

FEBRUARY 2005

and directional methods to locate CG lightning strikes to within 1 km. The data comprise the latitude, longitude, time, and other parameters for each strike and were in an ASCII format that could be read directly into ArcGIS. We calculated the average annual density of lighting strikes (units of km−2 yr−1) with the “density” function of the Spatial Analyst extension of ArcGIS (Fig. SB1), using a 1-km2 grid. We then compared the density map with terrain elevation from the U.S. Geological Survey’s National Elevation Dataset, obtained from the Pennsylvania Topographic and Geologic Survey. The elevation data were originally at a 30-m horizontal spacing and were resampled to a 1-km horizontal grid to be consistent with the strike-density data. Statistical analysis showed a weak, but statistically significant, negative correlation between terrain elevation and lightning strike density over southeastern Pennsylvania. Statistics calculated for individual counties showed either a weak, negative correlation or no correlation between lighting strike density and terrain elevation. The lack of positive correlation may indicate that orography is not the primary mechanism for thunderstorm formation over this region.

version utility has been updated to support negative elevation angles and theoretical beam patterns. As discussed by Olga Wilhelmi in this issue, the atmospheric science community has developed many of its own visualization and application standards [e.g., Grid Analysis and Display System (GraDs), Ferret, National Center for Atmospheric Research (NCAR) graphics]. The tendency of each community of users to build its own applications has the advantage of focusing on discipline-specific issues, but limits interactions with other research and applications communities. These problems are beginning to be addressed by broader adherence to community standards. I offer the following predictions to stimulate discussion. Now, remember that these are just forecasts, and I am a meteorologist. Perhaps they will prove more accurate than my 1994 prediction: 1) Pedigrees or “black belts” in meteorology will no longer be prerequisite for anyone to use and manipulate hydrometeorological data. The implementation of geospatial interoperability standards will dramatically increase the productivity of those individuals who presently spend 50-plus percent of their time writing code to convert data formats. 2) The weather community will eventually come up to speed in GIS education. No job description citing “GIS” skill currently exists in the NWS. (Traditional geography departments are also behind in GIS education, many reserving GIS topics for graduate study, even though Earth system science programs in secondary schools are using GIS.) 3) Web-enabled databases will transform how we use and distribute weather data—that is, unless we continue to embrace archaic data structures that require special-purpose algorithms and impede geospatial queries. 4) The Hierarchical Data Format version 5 (HDF5) will not supplant netCDF as the format of choice, but both will be challenged by a hybrid which combines features from each. HDF5 has been selected by the DoD to be a Joint Tactical Architecture standard. NetCDF has not. 5) GIS will play a central role in monitoring climate change. Evidence may be pointing toward land use as a large factor in climate change, perhaps dominating greenhouse explanations of tropospheric temperature rise in the last century. If this is so, GIS techniques will be needed to track landuse/land-cover changes on a global basis.

AMERICAN METEOROLOGICAL SOCIETY

FIG. 2. Theoretical beam coverage patterns for NEXRAD radars located in Washington State and Oregon. Beam blockage is depicted for terrain obstacles, where beam height is located within one vertical beam width of surface elevation. Graduated color (red) is used to show percentage beam blockage for cases of grazing incidence.

FOR FURTHER READING Pielke, R. A., Sr., G. Marland, R. A. Betts, T. N. Chase, J. L. Eastman, J. O. Niles, D. Niyogi, and S. Running, 2002: The influence of land-use change and landscape dynamics on the climate system: Relevance to climate change policy beyond the radiative effect of greenhouse gases. Phil. Trans. Roy. Soc. London A. Special Theme Issue, 360, 1705–1719. Reed, B., S. Shipley, J. Stauch, D. Gillespie, M.Walth, S. Young, C. Wang, J. Mulligan, J. Overton, J. Zajic, and A. Goldberg, 2004: Application of GIS for data quality monitoring in the NPOESS Ground Data Processing System. 20th International Conf. on Interactive Information Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, Seattle, WA, Amer. Meteor. Soc. Shipley, S. T., I. A. Graffman, and J. K. Ingram, 2000: GIS applications in climate and meteorology. ESRI International User Conf., San Diego, CA, ESRI. [Available at http://gis.esri.com/library/userconf/proc00/professional/papers/PAP159/p159.htm.] ——, ——, and R. E. Saffle, 2005: Weather radar terrain occultation modeling using GIS. 21st International Conf. on Interactive Information Processing Systems (IIPS) for Meteorology, Oceanography, and Hydrology, San Diego, CA, Amer. Meteor. Soc.

FEBRUARY 2005

| 173