WEB-8700 Weather Web Remote Observation System
General Description
The Weather Web™ Remote Observation System (WxWEB) provides emergency response teams with real time meteorological parameters from multiple, spatially-separated locations. It can be rapidly deployed and provides a robust and fully automated system for continuous unattended wind observations.
Data from the Weather Web can feed numerical weather prediction models such as HPAC to provide decision-makers with life-saving information about the direction of chem-bio plumes. Weather Web is based on networking and compatible with industry standards.
Real time data supplants the range of traditional human observations by automating manually intensive tasks. Weather Webs let personnel perform more important value-added tasks, and helps to eliminates subjectivity of human observers. In perilous situations, such as NBC attacks, it permits human personnel to leave affected areas for safer ground.
The Weather Web acquires data from remote RF-linked surface wind sets and optional Total Sky Imagers to obtain cloud data for real time wind profiles. The system automatically:
Checks status of TCP/IP network links
Acquires and displays real time wind data
Processes sky/cloud imager results
Integrates with HPAC runs
Features
Fully automated real time wind monitoring
Only commercially available solution for supporting HPAC with wind data
Professional-grade reliable met sensors
Standards based networking leverages existing IT investments
Requires minimal user training
Optional YESDAQ database supports post-event audits
Radio-MET Weather Post RF linked met set
Applications
Temporary or permanent observing system for homeland defense or tactical military applications
Numerical Weather Prediction Input for weather forecast models and
Plume Dispersion Model data input (HPAC)
Ground truth validation of remote sensing platforms
Spaceflight operations where the view to locations beyond the lower troposphere is critical to mission success
Scientific research on global warming and climate change
Labor reduction of human observers at cost-limited or remote/hostile locations
Keeping People Out of Harm’s Way
When a nuclear, biological or chemical (NBC) attack occurs, there are both short and long term impacts and casualties. In the initial first few minutes, precise knowledge of current and near-future winds is critical to optimally moving people out of the death zone. The YES Weather Web is a "rapid response" deployable wind measurement system that can be used drive numerical weather prediction forecasts (such as DTRA’s HPAC plume dispersion model). Output from these operational models can then be used to direct tactical and emergency response teams.
In a nuclear attack, the obvious blast damage occurs instantly is followed by a radioactive cloud of debris that reaches into the stratosphere. For nuclear attacks, upper air radiosonde soundings are necessary to provide vertical wind profiles to altitudes on the order of several dozens of km. Pressure, temperature Humidity and wind data conventionally are provided via Skew-T charts to give meteorologists key knowledge of the sate of the atmosphere and can be used to predict where fallout will occur. An automated upper air radiosonde system such as the YES Mobile ARL-9000 can provide this high altitude wind data, and it can be used to set initial conditions for operational mesoscale numerical weather prediction models such as COAMPS or MM5.
In contrast to nuclear events, chemical and biological attacks take place much closer to the ground, in the lower troposphere. Winds near the surface transport chemical or biological agents, and actually predicting what will happen to a plume in the boundary layer under stable conditions with a high degree of confidence remains an active research area by DTRA and others. The more stable the atmosphere is, the more difficult it becomes. The challenge here is to make accurate wind predictions with scarcity of data, and that is where the Weather Web enters the picture.
The Weather Web concept involves both a physical web of sensors geographically distributed about the combat/conflict area. It also provides relies on a virtual communications web of TCP/IP connections to both acquire sensor data and distribute it to decision makers in real time.
The Weather Web relies on SQL and Java software technologies to transparently provide data access to those who need it without requiring specialized software to reside on their local weather workstation. In times of disaster, there is simply no time to install or configure software. With the weather web, if a user is connected to the Internet backbone, they simply point a web browser at a URL. While the surface wind sensors are generally limited to a range of less than ten and require line of sight for their UHF 403 MHz links, access to it is limited only by the extent of the installed TCP/IP infrastructure.
At the core of the software is the Yankee Environmental Systems Data Acquisition system (YESDAQ), a powerful software system that automatically collects, archives and manages environmental sensor data from one or more surface observation sites. YESDAQ data can feed real time initial conditions for on line meteorological forecasting ingest systems such as LAPS. Once data are ingested and analysis fields are created, numerical weather prediction models can be run (such as MM5, ETA, RAMS, COAMPS, ARPS or WRF).
YESDAQ provides complete, unified remote instrumentation data management with support for optional Total Sky Imagers, Rotating Shadowband Spectroradiometers, LIDAR Ceilometers, Surface Meteorological Sensors and Upper Air Automated Radiosonde Launchers. Using TCP/IP networking, YESDAQ collects instrument data in real time from anywhere on the network, while simultaneously displaying real time data via web browser clients.
As the following figure shows, YESDAQ can be setup to support automatic remote data collection from multiple instrument sources, archive collected data in a relational database that is fully replicated across multiple servers. Remote visualization of archived data via web browsers in real time and the capability to link to third-party forecast and analysis tools via ODBC is possible.
Once remote sensor data is collected into YESDAQ, users can access it in real time via:
A built-in Web server that provides access through any Java-enabled Internet Web browser both as raw text and graphically via the DVE for each instrument type.
ODBC and JDBC drivers, which let you connect to the YESDAQ database and use third-party tools such as Matlab, Splus, IDL, or Crystal Reports to perform sophisticated analysis tasks (for example, cross-instrument data fusion, statistics, or linking to your own custom Java or MFC applications.)
YESDAQ is based upon MySQL, a licensed open-source technology that is highly reliable and well tested. Your data are not locked up in a proprietary system. When you direct your Web browser to data collected by a particular instrument, you are actually making live SQL requests into the database. Queries are then plotted via the instrument’s DVE component to let you wade through data in a variety of ways. Industry-standard ODBC and JDBC links let you use any number of third-party tools to perform specialized down stream data analysis, forecasting or sensor data fusion tasks. This flexible architecture lets you expand your data processing capabilities by adopting newer and better tools as software and networking technology improves. Other native programming interfaces are supported, including Perl.
A Look at Various Deployment Architectures
Because it is based on TCP/IP protocols, the system leverages the power of the Internet to both collect and distribute real time data from remote sites to users. Three basic topologies are supported: a single data type/instrument type with a single central server, mixed data types/instrument types with a central server, or multiple data types and several replicated servers. Managing any geographically distributed network requires careful planning and execution, as well as user training. In its simplest form, a metropolitan area can be covered by several dozen surface RadioMet sensors that are RF linked back to a central receiver and display workstation. However, more complex configurations are possible, using either different types of sensors (heterogeneous) or wider geographical coverage (homogeneous) topologies. The following shows a centrally managed wide area network of several surface sensor types.
The next figure shows a distributed but homogenous network of sensors of the same type, centrally managed.
The next figure shows a distributed military weather network collecting data from multiple instrument types and replicating data across multiple YESDAQ servers in the middle east, England and the US that are located close to end-user, downstream applications such as HPAC.
In this worldwide example, both Monterey CA, and Omaha NB forecast centers have YESDAQ hosts, with remote hosts at RAF Croughton in the UK managing a local Total Sky Imager and launcher, a host in Saudi Arabia managing a fleet of ARL launchers throughout the Middle East, and aboard a Navy aircraft carrier in the Pacific. Such a network can feed surface data in real time to existing weather prediction models and other no-casting forecast tools. Data fusion ranging from simple calculation of wind chill all the way to Total Sky Imager-based determination of winds at the cloud base are now possible.
The Weather Web’s underlying YESDAQ software has three primary software components: Service Manager, Database service, and Application/Web service.
Once sensor data is in the Weather Web’s YESDAQ repository, you use the Data Visualization Engine to browse it graphically. DVE components such as the Skew T chart viewer provide visual data browsing for each type of sensor/instrument, allowing for specialized data displays across multiple sensor and data types.
Core Wind Sensor Technologies
The Weather Web can be permanently installed or rapidly deployed by minimally-trained personnel. It feeds data via TCP/IP into back end HPAC plume dispersion tools and mesoscale models operating either in the field or at back end data centers. The weather web consists of multiple core wind sensor technologies: networks of battery powered, radio linked conventional surface anemometers setup as clusters (TMS-7200). Optional total sky imagers (TSI-880) and upper air Automated Radiosonde Launchers (ARL-9000) can be added at any time to facilitate better characterization of the upper atmosphere. Imaging and in situ radiosonde sensors can characterize winds through and beyond the Atmospheric Boundary Layer.
Real Time Information Dissemination
Rarely will you find a situation where you have all the responding players collocated in a chemical release scenario. In fact, location of weather forecasters, contaminant dispersion experts, and users of their predictions may be changing often during a chemical release response. Communications connectivity between players in such a dynamic environment must be rapidly re-configurable for all to maintain situational awareness. The USAF uses the Joint Battlespace Infosphere (JBI) offers the means for the needed rapidly re-configurable networks. The JBI is a high-level publish/subscribe software networking protocol which dynamically handles changes in user location. In this application, JBI integrates weather and contaminant dispersion experts with the users of their predictions in a C2 environment. The First Look Plume Analysis JBI client obtains the forecast and current observations from the Weather JBI client to select an estimated plume extent from a diverse set of pre-run Hazard Prediction Assessment Capability (HPAC) analyses for quick retrieval of plume prediction information closest to current conditions.
The initial response to a chemical weapon attack is when the decisions most affect the outcome and when there is the least amount of information on which to make these critical decisions. To be useful, information has to be accurate and must be delivered to the right people. When dealing with chemical weapon effects, weather forecasters and contaminant dispersion experts need to work cooperatively to develop an accurate forecast of where the chemicals will be transported and deposited over the next several hours. Because of the dynamic situation associated with a chemical attack, these two groups may not be collocated. Additionally, their locations may change during the crisis. This would necessitate a change in information flow to route information to new users who previously did not require it or to the same users at different locations. Once the dispersion forecast is available it needs to be quickly disseminated to many users with continual updates that accurately reflect the changing situation. The solution to overcoming these difficulties is found in a loosely coupled dynamically (re)configurable information system that provides connectivity between real time weather observations and a chemical hazard analysis capability.
The DoD’s Joint Battlespace Infosphere (JBI) provides the flexible conduit to dynamically handle these changes in information needs. The JBI, a high-level publish/subscribe software networking protocol, is used to integrate weather software and contaminant dispersion software in a C2 environment. It provides a flexible networking mechanism for maintaining the currency of the weather and chemical dispersion extent while keeping mobile responders and commanders in contact. A real-time estimate of chemical plume extent makes sense only if there are real-time weather observations to use as inputs to the analysis. The Weather Hazard Client has been developed to publish current weather observation to subscribers that are part of the JBI. The First Look Plume Analysis Client works in conjunction with the Weather Hazard Client to provide a first estimate of a chemical plume extent. This JBI Client uses chemical type, terrain, ground moisture and wind speed to choose the appropriate HPAC simulator analysis, from a diverse set of pre-run HPAC scenarios. This initial estimate will lack the accuracy of a detailed HPAC analysis but will supply the results immediately. This will be followed by a full HPAC simulation, which will take longer, but be more accurate.
Scenarios mentioned below include situations that could be equally applicable in both military and civilian domains. A military scenario might involve predicting the spread of contaminants in the air after a strike on a Weapons of Mass Destruction (WMD) Production Facility. A hazardous chemical plume prediction, in this case, is necessary in mission planning to (a) ensure minimum danger to noncombatants near the facility, and (b) prepare for the possibility of a Combat Search and Rescue (CSAR) mission into or near the contaminated area if the attacking aircraft is shot down near the plant and the pilot has to be extracted. This second point is of particular interest since a rescue helicopter crew’s ability to directly track visual cues (concerning its own security, non-radio signaling from the downed pilot, etc.) is considerably impaired by chemical suits. In a civilian context, an explosion inside a chemical plant could release toxic contaminants into the air, producing a plume so hazardous that people near the plant need to be evacuated by emergency response teams. This is analogous to a military unit coming under chemical attack requiring a rapid estimate of plume extent for evacuation planning.
To obtain the rapid first look forecast, parametric HPAC analyses are run that quantify the most significant input variables. The set of variables was chosen to represent a small collection of scenarios distinguished by one of several wind speeds, ground moisture conditions, chemical agent types, and terrain types. Since these analyses were prepared in advance of their intended use, the wind direction parameter in each file is fixed as originating from the north in all HPAC runs, but rotated to the true wind direction upon selection, as described below. The geographic location of interest, chemical agent type, terrain type, and ground moisture are published C2 intelligence and planning documents. With the geographic location defined the user can then subscribe to the correct localized weather published by the Weather client to obtain wind speed and direction. The First Look Plume Analysis client selects the data file most appropriate to the given conditions, and rotates the data about the geographic location to coincide with the given wind direction. The modified data file is then selected..
Default HPAC Parameters must be carefully considered. The choice of north as the default wind direction in the collection of typical situations represented by the HPAC runs is arbitrary, since the plume extent’s current orientation is determined from the wind direction in the subscribed weather data. Direction north was chosen simply because it is represented as 0 degrees angular displacement. By similar reasoning, the default location of the chemical incident in each of the typical case HPAC runs is 0 degrees longitude and 0 degrees latitude. The correct longitude and latitude of the WMD production facility is published in C2 plans. The adjustment of pre-run HPAC plume data for current conditions involves the angular rotation and geographic translation of the HPAC data, each item of which is another point value of concentration. The choice of location (0, 0) for the HPAC runs is motivated by more than just selection of default values. Since a rotation must be performed to orient the data to correspond to current wind direction, the rotation can be correctly done only at the coordinate system origin, followed by translation of the data to the location of the chemical incident. Arbitrary selection of any other chemical incident location for the HPAC runs would require an additional translation to (0, 0) before rotation and translation (in the opposite direction) again. The rotation-translation matrix formula takes into account the combining of two different systems for angular displacement, one system used by meteorologists for the direction from which prevailing wind originates, and the other used by mathematicians, consistent with classical definitions of trigonometric functions. Since the significance of the trigonometric functions is derived from a change in wind direction to account for current conditions, a clockwise displacement from north is transformed into the equivalent counterclockwise displacement from east.
The Weather Hazard Client functions as a local publisher of weather information. This Client automatically downloads imagery (GIF) and textual documents (html) from the Air Force Weather Agency at Offutt AFB on an hourly basis. The files are only retrieved, however, if they are more recent than the ones already obtained. These retrieved files are saved in a specified directory on the local file system of the Weather Hazard Client computer where they are available for publication to the JBI. To publish a particular weather image and/or airfield forecast/observation (html docs), the relevant set of XML metadata is selected using the Client’s GUI and the "Register" button is pressed. This notifies the lookup service that the Weather JBI Client has this product available. The "Publish" button completes the publication process. Once pressed, the user is prompted to select the directory where products associated with that metadata are stored. The associated product is then read in from that directory, wrapped as a JBI information object, and published every time a product of that type is downloaded from AFWA. Anyone with a JBI Client can subscribe to imagery and documents of interest and receive updates every time a new product is published.
For a chemical incident response, the First Look Plume Analysis Client subscribes to the airfield weather observation object for the nearest airfield to the site of the release. This observation includes the wind speed and direction information needed to make the proper selection of a representative plume from the pre-run HPAC data. By judicious choice of typical scenarios, pre-run hazard analysis and associated graphics can be used to quickly predict, in an offensive mode, the effect of a strike on a WMD production facility, or in defensive mode, to warn military units of the damage expected from an attack in progress.
REFERENCES
Specifications:
Met parameters: Wind Speed and Direction, Air temperature, % Relative Humidity (or dew point), Atmospheric Pressure; (optional alpha/gamma/beta radiation)
Wind Range: Wind speed: 0-60 m/s (134 mph)
Wind Gust survival: 100 m/s (220 mph)
Wind Direction Azimuth limit: 360° mechanical, 355° electrical (5° open)
Wind Speed Accuracy: ±0.5 m/s (1 mph)
Wind Direction Accuracy: ±5 degrees
Starting Threshold: Propeller: 1 m/s (2.2 mph)
Wind Direction Threshold: 1.7 m/s (3.8 MPH)
RF Telemetry
Link: Each station communicates cooperatively to a single base station
Frequency: operating in the 400-450 MHZ range
Modulation method: Frequency Modulation
Data format: Manchester encoding
Mechanical
Mounting: Directional antennas with bracket capable of handrail or wall mounting
Receiver: RS-232 interface, Windows Display
Software
HPAC interface: Conforms to DTRA requirements for real time wind data ingest
Optional Data Merge Facility
Managing data from multiple types of imaging and in situ sensors over an extended period produces a sizable data repository. In the weather web all data can be collected and stored for later display or further analysis in YESDAQ, a mySQL-based open source relational database with ODBC/JDBC connectivity to other downstream applications such as HPAC.
Yankee Environmental Systems is the only provider of this technology that integrates with HPAC for the emergency management community. Portions of this technology were developed and tested by the Army and Air Force for force protection applications and have been adapted for civilian use.
For more information, please contact YES technical sales.