Automatic Weather Stations

Sierra Leone Meteorological Agency has installed 8 new sufficiently sophisticated automatic weather stations (AWS) that now play the central role in the agency’s observing network. These AWS have been installed and maintained by Earth Networks through an approach being promoted in the UNDP’s Programme on CIRDA, which exploits the capabilities of the latest generation of smart, integrated, all-in-one (AIO) AWS, supplemented where necessary by even more powerful stand-alone data loggers, to provide sustainable observing networks for the 11 sub-Saharan African countries partnered with the support programme.

In these installed modern AWS, a specialized on-board computer or ‘data logger’ handles the collecting and initial processing of sensor readings. This has eliminated the need for human observers, handwritten observing forms, and calling or mailing in the recorded observations to a central office. Within a few seconds of being made, observations from AWS spread across a wide region can be collected at a central location, quality-checked using a consistent set of rules, archived, and made available for use by forecasters, climatologists and other users.In these all-in-one systems, the majority of the sensors and the related electronics are housed in one package. This significantly reduces the workload during deployment and subsequent field maintenance. These AWS exploit the cell-phone network to link the AWS to a central data collection facility. These cell-phone networks are subsequently used as a means to deliver hydromet services and information - another key component of this new vision.

The SLMD’s AWS are now ‘smart’, incorporating an on-board computer that autonomously provides for the generation and transmission of formatted meteorological reports, changing sampling rates and/ or special observations when preset environmental thresholds are crossed, and providing alert messages when preset thresholds in key variables are exceeded.

Following guidance from the Earth Network central data-collection facility, the distributed network of individual station computers manage all the necessary communications protocols. Individual stations have some storage capability, so that when communications to the central collection point are interrupted, data continues to be collected. They are then forwarded to the central collection point when communications are restored. This not only means that one is comparing apples to apples for long-term climate monitoring, in an ideal world, this could also mean automated storm, lightning and flood alerts for local communities.

The last decade has seen rapid advances in both sensor design and packaging technologies that has led to the recent emergence of a generation of ‘integrated, all-in-one’ automatic weather stations (AIO AWS). Typical AIO AWS contain sensors for measuring air temperature, relative humidity, precipitation intensity, precipitation type, precipitation quantity, air pressure, and/or wind direction and speed, all contained in a single package. Most moving parts have been eliminated, with wind speed and direction sensed by ultrasonic anemometers. Overall, AIO AWS prices are quite reasonable, particularly when long-term maintenance savings are considered. Most importantly, accumulating experience suggests that properly designed AIO AWS do quite well in difficult environments.

A baseline network of integrated compact automated weather and lightning sensors on mobile communications towers:

• Mobile communication towers provide optimal security, power, and communications
• Leveraging Earth Networks regional and global networks

This critical infrastructure provides foundation for:

• Real-time nowcasting and advance storm warnings
• Precipitation monitoring and accumulation estimates
• Forecasting on a variety of timescales

Earth Networks Total Lightning Sensor

• Total lightning detection (in-cloud and cloud-to-ground)
• Wideband electrical field recorder (1 Hz to 12 MHz)
• Records and transmits lightning flash waveforms
• Real-time data transmission
• Compatible with mounting on mobile telecom towers

Automatic Weather Station (AWS)

• Integrated AWS with no moving parts or maintenance
• Measures all weather parameters with high precision
• Real-time data transmission with hours of data storage
• Integrated with total lightning sensor for storm tracking

Earth Networks Network Appliance (ENNA)

• ENNA is a microprocessor-controlled computer
• Provides key connectivity, diagnostic, calibration and data archival functionality
• Ensures high quality data is transmitted from weather stations and lightning detection networks

Analysis and Alerting Services

Dangerous Thunderstorm Alerts (DTAs)

Dangerous Thunderstorm Alerts (DTAs) by Earth Networks provide advanced notification of the increased threat of severe weather moving into an identified area. A DTA alert is issued when there is a high frequency of lightning detected by the Earth Networks Total Lightning Network™ (ENTLN) indicating the increased potential for: lightning strikes, heavy rain rates, high winds and hail activity. The alert is updated every 15 minutes until the dangerous weather activity is no longer a threat and the alert expires. The advanced technology used within ENTLN enables the detection of both in-cloud and cloud-to ground lightning (otherwise known as total lightning). High rates of total lightning activity serve as precursory indicators of the potential for severe weather activity.

Earth Networks issues a Dangerous Thunderstorm Alert when the lightning detection rate exceeds high levels. These alerts are available through a data API that will return the alert information in CAP format. The alert CAP feed includes a polygon encompassing the area at risk, direction and speed of the severe lightning activity, cities in the route of the storm and current observations from weather stations near or in the affected area. A ready to use weather bulletin text is also provided within the CAP feed.

• Severe storms identified by monitoring lightning flash rates and rate changes
• DTA is 45 minute threat zone based on storm vector; re-analyzed every 15 minutes
• Early warning for: severe thunderstorms, high winds, hail storms, tornadoes, and cloud-to-ground lightning

SM Rainfall Monitoring and Estimating

• Radar imagery (dBz) using solely lightning activity data
• Visually identify and track severe weather and rainfall
• Monitor aggregate rainfall in real-time and monthly/annual for flood and drought warning
• Inexpensive (many times less!) alternative to radar with comparable imagery

ENcast Hourly Forecasts

The Earth Networks ENcast 15-Day Hourly Sensor Forecast Feed provides the MDAs and various stakeholders with a variety of hourly forecast variables up to 15 days in the future that will help organization make informed decisions. The ENcast 15-Day Hourly Sensor Forecast is available through Representational State Transfer (REST). This is an efficient and guaranteed delivery method utilizing web service (HTTP) explicitly where stakeholders can choose a file type, structure, and data variables from a menu of options. The forecast data, which is available through REST in JavaScript Object Notification (JSON) format is streamed into CIDMEWS-SL via the GeoEvent Server’s streaming and feature services functionality:

• Forecasts for two weeks out: updated hourly based on latest forecast model runs
• Hyper-Local: uses data from existing and new stations
• Fastest Updates: use of real-time AWS data or tuned to latitude/longitudes where no station data exists
• Lowest Forecast Error: accuracy and nowcast advantage
• Data hosting and delivery system ensures data quality control checks and sensor operability to deliver highest quality data at lowest maintenance cost
• Streaming total lightning detection (in-cloud and cloud-to-ground) data, recorded observations, and calculated weather data: 27 different variables
• Total lightning data is used instead of costly radar for storm cell identification, tracking, and alerting as well as real-time rainfall monitoring and estimating
• Hourly forecasts using ensemble of top global models (ECMWF, GFS, GEM, etc.) and weather and total lightning data to localize and enhance performance

Web-based display allows the CIDMEWS-SL user to view forecast information for specific locations (Bo, Kenema, Koidu, Mokaba, Rogberi, Kabala, Makeni and Wilberforce), timescales, and variables:

• Temperature
• 24hr High Temperature
• 24hr Low Temperature
• Wind Direction
• Wind Speed
• Dew Point
• Cloud Cover
• Thunderstorm Probability
• 1hr Precipitation Probability
• 1hr Accumulated Precipitation
• Fog Probability
• Visibility
• Rain Probability
• Surface Pressure
• Surface Insolation