ASCAT - Ocean satellite wind observations
|Provider:||EUMETSAT - European Organisation for the Exploitation of Meteorological Satellites|
|Resolution:||0.125°, 7.5nm, 13.9km|
|Model duration:||all data at time: 0 hr|
|Parameters:||wind, time at sample|
|GRIB model date:||Tue May 24 20:41 2022 UTC|
|Download date:||Tue May 24 21:52:45 2022 UTC|
|Download delay:||0hr 52min|
Note: the Download delay is the amount of time required for the GRIB model to compute its forecast and then for the LuckGrib cluster to download the data and make it available. The LuckGrib delay is generally less than 10 minutes, the remainder of the delay is the model compute time.
What is ASCAT?
The purpose of this model is to provide timely ocean wind observations. ASCAT wind data can act as a source of truth for what the actual wind was at a location and time on the ocean.
There is a large network of weather stations that provide real time measurements of the wind, however these stations are mainly confined to land areas. There is a sparse collection of buoys on the ocean which provide weather observations, however these buoys are largely concentrated in areas relatively close to shore. There are large areas of open ocean that do not have any buoy’s present, and finding wind observations for those areas can be difficult.
There are currently (April, 2022) two satellites orbiting the earth, generating ASCAT data:
The following description has been taken directly from the NOAA Office of Satellite and Product Operations (OSPO) product page:
The Advanced Scatterometer (ASCAT) winds products are processed by NOAA/NESDIS utilizing measurements from the scatterometer instrument aboard the EUMETSAT Metop satellites. The instrument uses radar to measure backscatter to determine speed and direction of winds over the surface of the oceans. …
ASCAT wind data provides statellite observations of winds at the ocean surface.
Why use ASCAT data?
The ASCAT data are observations not forecasts, they provide a source of truth for the actual ocean winds.
The main purpose of this model is to compare its wind data to surface wind forecasts, to evaluate their accuracy. If a forecast compares well with the ASCAT data, then that forecast was skillful. Where a forecast and the ASCAT data differ, the ASCAT data should be trusted.
Most of the global wind forecast models are run on the synoptic times: 00Z, 06Z, 12Z or 18Z. This ASCAT model is biased toward making validation at the synoptic times easier.
As the ASCAT data is based on actual observations, it is always historic. By the time we have access to the data, it records something that happened in the past. This is also true of the 0H forecast data in a global weather model. The 0H forecast is always historic with respect to when the forecast data becomes available. Comparing the 0H forecast winds to ASCAT winds may provide insight into how well the forecast may perform.
LuckGrib ASCAT data.
There were many alternatives available in deciding how to present this data in GRIB form. This section is meant to explain some of those choices.
The ASCAT satellites are in a sun-synchronous orbit. Here is an example of a single orbit from an ASCAT satellite:
This data starts around north western Africa. That is the earliest (oldest) data in this image. The satellite then heads north to the pole on an ascending orbit. After crossing the north pole, the satellite descends and passes through the Bering Sea on its path to the south pole, and then heads north to the equator again.
One of the traits of a sun-synchronous orbit is that a satellite will cross the equator at the same local time and orientation to the sun. So when the satellite crosses the equator for the second time in this image, the sun is in the same relative orientation to the earth as it was for the previous orbit.
The reason for the slant in the data swaths is that the earth is rotating beneath the satellite. After a little study, you can recognize which data is from an ascending or descending swath by its orientation. Each orbit takes roughly 1hr 41min.
During the development of this version of the ASCAT model, one of the first experiments on how to package it for use by users was to present single orbits of data in the GRIB file, with the two satellites each in their own record. This presentation ended up being quite difficult to use in practice.
After a series of experiments, the final form for this model was chosen:
There are four GRIB records created for each day, each containing 6 hours of wind observation data. As the main purpose of this model is for the validation of forecast data at the synoptic times, the GRIB records created contain wind observation data which is centered on each synoptic time. Each GRIB record contains 6 hours of ascending and descending wind observations, from the metop-b and metop-c satellites.
GRIB @ 09Z: 03Z +++ 09Z GRIB @ 15Z: | 09Z +++ 15Z GRIB @ 21Z: | | 15Z +++ 21Z GRIB @ 03Z: | | | 21Z +++ 03Z | | | | hours: 00Z 03Z 06Z 09Z 12Z 15Z 18Z 21Z 00Z 03Z ...
For example, a GRIB record with a 15Z reference time may contain data from 09Z through to 15Z. This makes the 15Z data ideal for validating the 12Z forecast.
Read that sentence above again, as its crucial in using this data. If you want to validate GRIB data at one of the synoptic times, choose the ASCAT data 3 hours further into the future, and then that record will contain 6 hours of data, centered on the synoptic time you are interested in.
Of course, there may not be ASCAT wind data available in your area of interest at the forecast time, but you can start there and then search to find the freshest data for your location.
Note that each area in the world has an opportunity for a new update of data roughly every 12 hours. While there are two ASCAT satellites, they are closely aligned with each other. There is probably a good sciencey reason for this which escapes me…
Lastly, note that you may be interested in an area which is missed by the ASCAT data swath for any particular orbit. In this case, you can either look for older data which does pass over your region, or attempt to interpolate from data if it appears relatively close by.
Time at each sample
In addition to the model providing wind data, it also provides the time at each sample. In particular, it provides the number of seconds prior to the reference time. Using this data, it is possible to determine the time for each wind sample.
LuckGrib will display the valid time for each wind sample, as text, if and only if the GRIB time is exactly on one of the GRIB record times. When working with ASCAT data, you should ensure that as you alter the time in the app, that you use the app’s ability to move to the previous or next GRIB time (left / right arrow keys on Mac, central radial menu next / prev commands on iPhone and iPad.)
When this is done, you will see an information display such as:
If you do not see the line showing the valid time then either the file does not contain the time parameter, or the current GRIB time is not placed exactly on one of the synoptic times.
When comparing forecast data to ASCAT observations, you should be aware of the time you are looking for, and the time at each ASCAT sample.
Its possible to view a visual representation of the time contained in an ASCAT model. For example:
There is no particular meaning to the color - red is no better or worse than green, blue, etc. This image can be used to identify edges of ASCAT swaths. Time will vary slowly within a swath, following the orbit of the satellite. When you move from one swath to another, the valid time may jump quite a lot. For example, this area of the ASCAT image contains many overlapping swaths, as shown by all the differnent colors. Its just something to be aware of as you work with this data.
Timeliness of the data.
(The credits section below describes the cast of characters mentioned here.)
I believe the order of operations is something along the lines of:
- EUMETSAT / KNMI receives raw data from the satellites, processes it and delivers it in 3 minute chunks called granules. There is a lot of science magic accomplished in this step.
- PO.DAAC gathers these granules and combines them into orbits. A full orbit consists of the first granule that is ascending above the equator. An orbit is finished when the equator is encountered again. PO.DAAC makes these orbit files available quickly after the last granule arrives.
When a new orbit file is available, LuckGrib downloads its content and processes the most recent (approx) two days of data into a fresh model run.
If you use LuckGrib to download ASCAT data for a large area, frequently, over the period of 6 to 12 hours, you will see the final GRIB record slowly accumulate more and more orbit data.
The ASCAT model on the LuckGrib server updates frequently, but on a somewhat irregular schedule, ranging from 40 minutes to a little over an hour. The updates occur quickly after the data is available.
The primary source for the ASCAT data is EUMETSAT:
The Advanced SCATterometer (ASCAT) is one of the instruments carried on-board the Meteorological Operational (Metop) polar satellites launched by the European Space Agency (ESA) and operated by the EUropean organisation for the exploitation of METeorological SATellites (EUMETSAT).
The Royal Netherlands Meteorological Institute (KNMI) is involved in its processing:
KNMI is involved in the OSI SAF and the EUMETSAT Advanced Retransmission Service (EARS) ASCAT service as the centre where the Level 1b to Level 2 wind processing is carried out.
When reading material related to this data, you may come across references to EUMETSAT/OSI. OSI is the Ocean and Sea Ice department at EUMETSAT. There are also references to SAF/KNMI. SAF is the Satellite Application Facility at KNMI.
The raw ASCAT data is being obtained by LuckGrib through the Physical Oceanography Distributed Active Archive Center (PO.DAAC) at the Jet Propulsion Labratory (JPL).
The ASCAT data at PO.DAAC internally contains the following credit string: ‘EUMETSAT/OSI SAF/KNMI.’
PO.DAAC has the following data use policy:
Data hosted by the PO.DAAC is openly shared, without restriction, in accordance with NASA’s Earth Science program Data and Information Policy.