The Florida MesoNetwork:
High Resolution Meteorological Observations in Florida(1)
Charles Paxton and Andrew Nash
National Weather Service, Tampa Bay Area, Florida
A. Florida Weather
Florida has a diverse, weather sensitive population. Tourism, agriculture, and industrial businesses rely on weather information. Florida's geography plays a major role in weather patterns over the State. Large scale weather features such as cold fronts and tropical storms, cause varying conditions across the length of the State from Key West north to Jacksonville and west to Pensacola. The Gulf of Mexico and the Atlantic Ocean and Florida's numerous lakes generate sea, land, and lake breezes that also lead to considerable weather variations over short distances. Thus, frequent weather observations are important to indicate current conditions at points across the state and serve as an initial step for forecasts and as verification for previous forecasts.
B. Weather Needs
Florida's weather problems and needs come in a variety of forms. Occasionally, winter weather dips deep into Florida. Freezes are a significant problem for Florida residents, and in 1989 freezing rain and snow stranded tourists on I-75. The March 1993 "Storm of the Century" caused widespread destruction and many deaths from tornadoes, strong wind, and coastal flooding. An accurate assessment of off-shore winds and wind-driven tides near the coast could have saved lives and reduced property damage during that storm. Hurricane Andrew devastated south Florida, and brushes with other economy impacting tropical storms are common. Other threats each year are periods of drought and flooding making water assessment and management necessary. Hazardous material spills occur frequently, but accurate evaluation of weather conditions can mitigate the damage. Current weather condition appraisal is necessary for weather prediction for Florida Department of Forestry prescribed burns. Search and rescue operations are weather sensitive. Agricultural spraying regulations are dictated in part by weather conditions. Red tide outbreaks and movement are dictated by weather. Law enforcement agencies use weather data for reconstruction of cases. Current weather information is important for a wide variety of applications.
C. Mesonet Definition
The term mesonet incorporates two terms: mesoscale, which describes occurrences with scales on the order of tens of miles and tens of minutes or less, and network, which ties many independent units together. The data collected vary from site to site but some of the fundamental fields are wind, temperature, dew point, humidity, rainfall, and pressure. The cost of installing and maintaining a mesonet is high. Each land based site costs $5000-$15000 depending on instrumentation and quality. Buoy-based sites are costlier to install and maintain, ranging in price from about $30,000-$80,000.
D. Other Mesonets
A premier mesonet in the United States is the Oklahoma Mesonet. This is a collaborative project between the University of Oklahoma and Oklahoma State University and was installed as a new system during the early 1990s using $2 million gained from oil-overcharge funds that were available through the U.S. Department of Energy. The two universities also provided another $0.7 million to support the design, implementation, and initial operation of the mesonet. The Oklahoma Mesonet consists of 111 automated observing stations (at least one site per county) that continuously monitor numerous important weather and soil parameters. The Oklahoma Mesonet provides for automatic data collection every 5 minutes, data sharing every 15 minutes, and a sophisticated data transfer network. Real-time data are provided as a service to registered data subscribers of the Oklahoma Mesonet Project. A variety of products and climate summaries are produced from the data. Other mesonets are being developed in Utah, California, Texas, Washington and Colorado.
A. The Florida MesoNetwork Concept: Use Existing Sources
Automated weather observation sites have been installed by many different federal, state and private agencies throughout Florida. Our state has a fairly high density of surface observations, but most networks have been set up independently of the others. Joining the existing smaller networks into one large cohesive network allows all of the contributors access to much more data than they now have. Numerous weather observation sites exist over the peninsula but observations are lacking over the Gulf of Mexico and Atlantic Ocean. Once the existing observations are networked, new automated observation sites including buoys may be placed at geographically significant sites requiring coverage.
Frequency of observations is important also. Observations are collected at least hourly, but finer time resolutions should be strived for. To complete the data set, twice daily data from National Weather Service radiosonde sites lend a three dimensional view of the atmosphere. This enhanced data set provides various users a much more detailed picture of the weather. The Florida MesoNetwork is supported, in part, by the University of South Florida Department of Marine Science, the University of Florida Institute of Food and Agricultural Sciences, and the Florida State University Department of Meteorology. The staff at the NWS office in Ruskin has worked to consolidate existing data into the mesonetwork described below.
The primary benefits of incorporating existing smaller networks over building a system from the ground up are: low initial cost and maintenance performed by equipment owners. A high density network also allows comparison of nearby observations which may reveal data quality inconsistencies of a particular site. Data contributors receive more data than they put in. Benefits from a mesonet are improvement of atmosphere assessment leading to an improvement of forecasts.
A dense network of sources with more than 600 observation sites is available in the Florida MesoNetwork. Instrumentation at various observation sites varies from a single measurement such as rainfall to a wide array of sensors such as multiple anemometers for wind measurements at different levels. Observations taken in support of aircraft observations have visibility and cloud coverage information. Coastal and offshore sites may have wave or water level and temperature measuring devices. Agencies with automated weather observation data are show in Table 1:
Table 1. Observation Sources
|Source||Number of Sites|
|National Weather Service/ Federal Aviation Administration / Department of Defense||75|
|National Ocean Service/ National Oceanographic and Atmospheric Administration||25|
|U.S. Forest Service||18|
|Florida Water Management District (mostly rain gauges)||300+|
|United States Geological Survey (mostly rain gauges)||94|
|Army Corps of Engineers||5|
|University of Florida, Institute of Food and Agricultural Sciences, Lake County||5|
|University of South Florida, Department of Marine Science||10|
|Tampa Bay Physical Oceanographic Realtime System||7|
|National Aeronautics and Space Administration||20|
|Tampa Electric Company||5|
|Sugar Cane League||4|
Communication with observation sites is the most complex part of the Florida MesoNetwork. Communication is through a combination of 1) radio telemetry, 2) satellite telemetry, 3) land line with modem, 4) Internet, or 5) standard NWS communication lines. Various observation equipment manufacturers have different protocols for communicating with field sites. A Pentium PC with four modems at the Ruskin NWS weather office is connected to operational lines and the Internet. The NWS Florida MesoNetwork Data Collection System collects and disseminates observation at least hourly. The data are available through NWS communication equipment and the Internet. Data storage and acquisition is important for post-event analyses and data are locally archived.
Access to the Ruskin NWS office home page via Internet (www.marine.usf.edu/nws/) allows users to obtain data from the Florida MesoNetwork in a variety of ways.
E. Data Format
Once collected, the raw data are decoded and arranged in a standard Florida MesoNetwork alphanumeric format. Next, the data are quality controlled by comparing particular parameters (temperature, dew point, humidity, sea level pressure and altimeter setting) by geographical region. Values for those parameters may vary considerably over short distances, therefore, the quality control program eliminates only the extreme outliers. Finally, the data are arranged in columns in the format shown in Table 2. The following units are used: latitude and longitude (degrees, tenths and hundredths), wind direction (degrees), wind speed and gust (knots), weather (codes are: T=thunderstorm, R=rain, RW=rain shower, S=snow, SW=snow shower, A=hail, H=haze, F=fog, - or + indicates light or strong intensity), visibility (statute miles), temperatures (Fahrenheit), sea level pressure (millibars), altimeter pressure (inches), wave height (feet), wave period (seconds), relative humidity (up to 99 percent). Rainfall (in inches) is included in a separate product.
Table 2. Observation format
MMDDHH IDIDI LATT LONG DDFF GGG OBOBO VVV.V SKY TTT DDD SLP ALT TW WV PD RH
110109 NSE 3070 8705 0000 F 5.0 SCT 57 55 150 999 93
110109 PNS 3050 8720 0000 R- 7.0 CLR 61 61 999 99
110109 VPS 3050 8650 3303 017 7.0 FEW 57 55 156 999 93
110109 PAM 3015 8560 1102 7.0 CLR 64 63 159 000 97
110109 42036 2850 8350 2717 034 76 74 112 989 77 12 08 93
F. Graphical Products
Data are transformed into several graphical formats for easier interpretation. GEMPAK and Interactive Data Language (IDL) analysis software packages plot, contour and graph the data.
Figure 1 is an example of an individual station plot with an east-southeast wind at 15 knots. Plots of temperature, dew point, pressure, wind direction and speed, and water height (wave or tide) are overlaid on a Florida map background (Fig. 2). Several zoomed sectors are available. Wind direction is indicated by a line oriented by compass location, and wind speed is indicated in increments by short attachments (5 knot), long attachments (10 knot) or a pennant (50 knot) to the wind direction line.
Plots show several fields on one map background while contour analyses are produced for a single field. Contour analyses are produced for temperature, dew point, relative humidity, and pressure fields (Fig. 3).
Time Series Graphs
Time series graphs (Fig. 4) show multiple fields for one location over a 24-hour period. The locations are selected from either a clickable map or an HTML form. The graphs are produced "on the fly" by IDL software. The vertical axis scales slide to accommodate any 24-hour range of temperatures.
3. Results and Discussion
The primary goal for the future is to increase the number of available sites and frequency of data collection. Equally important are derived alphanumeric and graphic climatological products in a visual climate data base. An example is a running total of precipitation on 30, 60, 90, 180 and 365 day bases, and accompanying graphical rainfall plots and time series. Other future graphics include:
2) Wave and water height contours and time series
3) Wind vectors and streamlines
4) Derived products, such as surface moisture convergence
5) Products derived from combined surface and upper level radiosonde data, thereby alerting forecasters to important changes through the atmosphere.
Another goal is to input the Florida MesoNetwork data into computer models. An example is the RAMS (Regional Atmospheric Modeling System) now run at the NWS Ruskin office. Finally, another archive site will be collaborated with the State Climatologist at Florida State University in Tallahassee.
The Florida MesoNetwork developed by the National Weather Service Office in Ruskin, is a low cost, high return weather observation system, in collaboration with many federal, state and private entities. The system provides current weather information for a variety of uses including agriculture, emergency management, and for National Weather Service model input and forecast operations. Unlike the 111 observation sites of the Oklahoma Mesonet, for example, built from the ground up at a cost of $2.7 million, the Florida MesoNetwork collects data from a wide variety of smaller, existing networks to provide more than 600 observations.
Communication is provided by a combination of radio and satellite telemetry, modem, Internet, and through standard NWS communication lines. Data are quality controlled by geographical region and put into a tabular format. A variety of graphical products is produced to provide easier interpretation. Graphical products include station plots and contours on maps and time series graphs for individual stations. Data are available through NWS equipment and the Internet and are locally archived. Anticipated improvements include increasing temporal and spatial resolution, visual climate data base, overlay of data on satellite imagery, a variety of derived products, and input of the data into a high resolution forecast model.