Home Page - Hydrometeorological Automated Data System

What is HADS?

The Hydrometeorological Automated Data Systems (HADS) is a real-time system for acquiring, processing, and distributing hydrometeorological data. It operates within the Meteorological Assimilation Data Ingest System (MADIS), managed by NCEP Central Operations (NCO). HADS primarily supports National Weather Service (NWS) programs, including Flood and Flash Flood Warning efforts conducted by Weather Service Forecast Offices and River Forecast Centers across the U.S. Additionally, HADS data enhances fire weather support, precipitation analysis, hydrological and meteorological modeling, and the verification of NEXRAD precipitation estimates.

HADS collects raw hydrological and meteorological observations from Geostationary Operational Environmental Satellites (GOES) Data Collection Platforms (DCPs). These platforms are managed by over 200 cooperators, including agencies like the U.S. Geological Survey (USGS), U.S. Army Corps of Engineers, Tennessee Valley Authority, Bureau of Land Management, U.S. Forest Service, Bureau of Reclamation, and various state and local natural resource departments. In return, the NWS shares hydrological and meteorological products with these agencies.

The National Environmental Satellite, Data Information Service (NESDIS), a NOAA agency, operates and maintains the GOES Data Collection System (DCS) at Wallops Island, Virginia. This facility downlinks satellite data, relaying it to HADS and the broader DCS community. HADS receives this data through the Local Readout Ground Station (LRGS) software and a dedicated file transfer circuit connecting to the National Weather Service Telecommunication Gateway (NWSTG).

Using LRGS as its primary feed, HADS continuously receives data from Wallops Island and the USGS EDDN every 5-8 seconds, buffering and processing it in two-minute cycles. This process includes translating raw data into Standard Hydrometeorological Exchange Format (SHEF) products. Given the complexity of this translation, HADS requires extensive site-specific information for over 18,500 locations.

Within the HADS database are metadata about each of the data sites, some of which are used to translate the data to SHEF. This includes the following:

Owner of the DCP
NESDIS Identifier
NWS Location Identifier (NWSLI) associated with the NESDIS Identifier
Decoding scheme
Type of data observed (river stage, precipitation, air temperature, etc)
Units of measurement for each value
Numerical precision of these values
Data interval for each type of data observed
Time and interval at which the DCP uplinks to a GOES satellite
Geographical location (location name, latitude, longitude)

During the processing and translation of the data from raw to SHEF form, the data are put into individual user reports, which are tailored for the needs of each of the NWS Weather Forecast Offices and River Forecast Centers. The data products are disseminated through the NWSTG which then routes the data products to the AWIPS Network Control Facility (NCF) which in turn broadcasts the data products on the Satellite Broadcast Network. The end users can downlink the reports via NOAAPORT. Additionally HADS will place all users data products on ftp server (ftp://madis-data.ncep.noaa.gov/hads/shef_products) where they remain a number of hours. Finally, HADS makes DCP metadata and decoded data available to the field and public via the HADS website.

HADS has many built in redundancies for data acquistion, processing, and dissemination to prevent outages.

DCP Characteristics

The GOES Data Collection Platforms operated and maintained by the cooperators have 4 basic components.

Datalogger for recording the sensor's information
UHF radio transmitter
Yagi or omni-directional antenna
One to many Environmental sensors

DCPs are capable of operating into two distinct modes, self-timed and random.

In self-timed mode, a DCP uplinks its data on an assigned channel and time. It also has an assigned transmission interval. The vast majority of DCPs transmit on a one hour cycle, but some transmit at 12, 15, 30 minute, 2 hour, 3 hour, and 4 hour cycles. Within each DCP's message, the actual interval of the data may be 1, 2, 3, 5, 6, 10, 15, 20, 30 or 60 minutes.

In random mode, also known as critical mode, a DCP will uplink a short message containing 1, 2 or 3 values of one or two 'critical' sensors. The threshold for this type of data transmission is dependent upon how the DCP has been programmed. Typically a random message is generated when a water level reaches and exceeds a predefined height or increases at a predefined rate. Random messages of precipitation data are typically generated when the rainfall rate for a defined time interval is met or exceeded.

Page Updated: April 8, 2025