ASCAT Version 2.1 Data Release (April 2016)
ASCAT data have been fully reprocessed as Version 2.1. The reprocessing was performed in order to implement a new Geophysical Model Function, C-2015, that was developed at RSS in order to reach a high level of consistency between all scatterometers and radiometer winds, at all wind speed regimes. The wind retrieval algorithm used for V2.1 is the same as in V1.2, with the exception of small changes in the quality control, rain-flagging, and ambiguity removal algorithm. The new V2.1 winds were produced using the recalibrated L1B sigma0 data (sigma0 Climate Data Record) from EUMETSAT. Please read the ASCAT Product Update Notes for more information.
The first Advanced Scatterometer (ASCAT) was launched on the EUMETSAT MetOp-A satellite in October 2006. It became fully operational in May 2007 and continues to operate today. Another ASCAT instrument became operational on MetOp-B when launched in September 2012. Both satellites, MetOp-A and MetOp-B carry identical ASCAT instruments. The main objective of ASCAT is the measurement of wind speed and direction over the oceans, though ASCAT is also used for studying polar ice, soil moisture and vegetation.
ASCAT is a C-band scatterometer with 3 vertically polarized antennas transmitting pulses at 5.255 GHz. The fan-beam antennae are oriented at 45, 90, and 135 degrees with respect to the satellite track. The antennae extend on either side of the instrument, resulting in a double swath of observations, each about 500 Km wide separated by a gap of about 360 Km. More information on the ASCAT instrument is available from the EUMETSAT web site and Figa-Saldana et al, 2002 (see references below). Scatterometers are essentially radars that transmit microwave pulses down to the Earth's surface and then measure the power that is returned back to the instrument. This "backscattered" power is related to surface roughness. For water surfaces, the surface roughness is highly correlated with the near-surface wind speed and direction. Hence, wind speed and direction are retrieved from measurements of the scatterometer's backscattered power.
Climate Quality Ocean Vector Wind Retrievals
Our long-term vision at RSS is to develop multi-decadal, highly accurate and carefully intercalibrated satellite data sets suitable for climate research (Climate Data Records: CDRs). This has already been achieved for the surface wind speeds from satellite radiometers, mainly by using a common Radiative Transfer Model for all the different sensors.
The development of a CDR is underway for the ocean vector winds from scatterometer measurements, which started in 1991 with the European ERS-1. The longest operating instrument so far has been NASA’s QuikSCAT, which was launched in July 1999. After the demise of QuikSCAT in November 2009, the European ASCAT scatterometer assumed a vital role in the long and continuous time series of surface ocean vector winds suitable for climate research.
The ideal approach to combining winds from multiple, different sensors is to develop a similar methodology and calibration standard for use in data processing. We have done that for both QuikSCAT and ASCAT. Operational ASCAT data are available from EUMETSAT OSI-SAF. The RSS ASCAT data are for climate consistency with QuikSCAT.
RSS ASCAT C-2015
We recently update the Geophysical Model Function (GMF) used to derive Wind Speed and Direction at 10-meter height over ice-free oceans from backscattered scatterometer signals. Over many years, RSS has developed and improved a Ku-band GMF. The most recent Ku-band GMF (Ku-2011) was designed for consistency with radiometer winds at all wind regimes, and used several years of backscatter ratios collocated with rain-free WindSat wind retrievals. We have applied the same development methodology to the C-band scatterometer data and the result, C-2015 GMF, has been used to process MetOp-A ASCAT data and produce Version 2.1 winds.
The quality of a GMF depends on: i) The quality control of the original backscattered signal; ii) The number of collocations between the observed sigma0 and a ground truth for wind speed, wind direction, and for each ASCAT radar beam incidence angle; and iii) The choice of ground truth wind speed data - aiming for high quality and accuracy at all wind speeds with a narrow time-colocation window with ASCAT. For our first processing of ASCAT (V1), we used SSMIS wind speeds as our ground truth. For the C-2015 GMF we used TMI and GMI wind speeds. The TMI data are available for most of the ASCAT mission and GMI is available during the later period. Both can be closely collocated to the ASCAT data. Rain rate measurements from TMI and GMI are used to select only rain-free ASCAT backscatter measurements for use in developing the C-2015 GMF.
The first version of ASCAT data, V1, was released in March 2014. We followed this with an update, V1.2, in August 2015. While there are several other ASCAT data sets in use today, our goal has been to release a set of ASCAT winds that are carefully intercalibrated with QuikSCAT and the radiometer wind speeds from Remote Sensing Systems, allowing for these ASCAT winds to be confidently used in climate study. Our work on the C-2015 GMF has produced climate-quality winds. We refer to this data set as V2.1 C-2015 ASCAT wind products.
The fully reprocessed V2.1 ASCAT data set, released in April 2016, extends from 2007 to the present. Processing is completed in an on-going fashion with approximately a 2-week delay. This updated data set brings ASCAT data into climate quality (0.1 m/s) with all other scatterometer and radiometer data available from Remote Sensing Systems. We use ASCAT L1B recalibrated radar backscatter values (sigma0) in our processing to obtain ASCAT winds. The difference between V2.1 and V1.2 winds are on the order of 5-6% higher at winds above 30 m/s. Other differences are discussed in greater detail in the ASCAT Update Report and comparisons of ASCAT V2.1 winds with those from moored buoys, QuikSCAT and satellite radiometers are also included.
C-2015 has also been used to process ASCAT-B and ASCAT-C.
We obtain 12.5 Km ASCAT L1B backscatter data from EUMETSAT. We have data for March 2007- the present with about a 2-week delay.
We reorganize the data into orbits starting at the south pole and process each into ocean winds using the C-2015 GMF. The methodology is similar to that of QuikSCAT for consistency. Two presentations, one on the ASCAT C-2015 GMF and one presenting the ASCAT wind validation results contain more information. The recalibrated radar backscatter values available from EUMETSAT are discussed in a presentation by J. Figa (see references below).
The impact of rain on wind retrieval is less severe for ASCAT C-band data than for QuikSCAT Ku-band data. Rain impact on scatterometer retrievals generally produces a positive bias at low wind speeds (due to signal backscatter by rain drops) and a negative bias at high wind speeds (due to atmospheric attenuation of the signal). The biases can be quite large for Ku-band scatterometers, but are much less for ASCAT, though not negligible and mostly limited to low wind effects. Unlike QuikSCAT, ASCAT wind retrievals in rain at high winds, such as tropical storms, are quite good. ASCAT winds displayed on this website and provided in the data files on our ftp server exist in both rain-free and rainy areas. When possible, we provide a rain-rate estimate from one of our radiometers in addition to a stand-alone scatterometer rain flag. We advise users to rely on both the scatterometer rain flag and the radiometer rain information within the file to remove any rain-contaminated data from their data processing as they see fit. See the notes within the ASCAT read routines for more information on how to access and apply the rain flags.
For the case of ASCAT, we use satellite microwave radiometers (TMI, GMI, AMSR2, WindSat) to determine if rain is present within 3 hours at the location of the ASCAT observation. In addition, all available radiometers are used to detect sea ice. Using the daily observations of sea ice, the scatterometer observations can be properly flagged so that reliable wind vectors can be obtained immediately next to the marginal ice zone.
The ASCAT data we provide have a similar format to QuikScat V4 and SeaWinds V3a data, however the resolution of the swath data (L2B) is different and one additional parameter has been added to the daily gridded maps. More details are provided below.
RSS ASCAT Scatterometer Data Products
ASCAT data products are in binary file format. Two types of products are available: 1 swath product which contains the data in one swath per file organized by swath rows perpendicular to the satellite path or 2) gridded binary data files similar in many ways to those provided for RSS radiometer data.
|Surface Wind Speed||Equivalent Neutral Wind speeds 10 meters above the water surface, derived from surface roughness (wind stress), roughly equivalent to an 8 - 10 minute mean surface wind.|
|Surface Wind Direction||The angle of air movement, north = 0 degrees.|
|Rain Info||ASCAT data can be less accurate in heavy rain, especially at low winds. Data products include a scatterometer derived rain flag, and co-located radiometer rain rates (mm/hr).|
|SOS map||Sum-of-Squares map identifying areas where the data poorly match the theoretical backscatter model GMF.|
Gridded Binary Files
We produce daily and time averaged (3-day, weekly, monthly) gridded data files by mapping the scatterometer orbital data to a 0.25 deg longitude by 0.25 deg latitude Earth grid.
Gridded ASCAT data are publicly available via FTP at: ftp://ftp.remss.com/ascat/metopa/bmaps_v02.1
Folders and file names follow these conventions:
|Time||Directory Path||File Name|
Where [year], [month], "yyyy", "mm", "dd", and "v" stand for:
|[year]||year folder||y2007, y2008 etc.|
|[month]||month folder||m01 (Jan), m02 (Feb) etc.|
|yyyy||year||2007, 2008 etc.|
|mm||month||01 (Jan), 02 (Feb) etc.|
|dd||day||01, 02, ... 31|
Note that 3-day and weekly files are named for the day they end on (including that 3rd or 7th day).
Data are encoded in single byte values. Each data file contains a sequence of byte maps, or bmaps, each representing the Earth at quarter degree resolution: 1440 x 720 bytes.
Daily files are byte arrays of size 1440 x 720 x 5 x 2 (longitude, latitude, parameter, orbit segment (ascending or descending passes)). The 5 parameters are: UTC Time of Observation, Ocean Surface Wind Speed, Ocean Surface Wind Direction, a Rain Flag / Collocated Radiometer Rain combination value, and the Sum-of-Squares (SOSmap). Two maps exist for each parameter: one of morning orbit segments (descending passes) and the other of local evening orbit segments (ascending passes).
|1440||longitude||0 to 360|
|720||latitude||-90 to 90|
|5||parameter||UTC time, wind speed, wind direction, rain info, sosmap|
|2||orbital segment||local morning (descending) passes, local evening (ascending) passes|
Thus, daily files contain 10 global maps:
[morning time, morning speed, morning direction, morning rain info, morning sosmap, evening time, evening speed, evening direction, evening rain info, evening sosmap]
Time averaged files (3-day, weekly, monthly) are byte arrays of size 1440 x 720 x 4 (longitude, latitude, parameter). The 4 parameters are: Ocean Surface Wind Speed, Ocean Surface Wind Direction, Rain Flag / Collocated Radiometer Rain Rate combination value, and the SOSmap.
|1440||longitude||0 to 360|
|720||latitude||-90 to 90|
|4||parameter||wind speed, wind direction, rain info, sosmap|
Time averaged files contain 4 global maps: [wind speed, wind direction, rain info, sosmap]
For time-averaged scatterometer data, wind speeds are scalar averaged, while wind directions are vector averaged. Thus, if daily observations record strong winds blowing in opposing directions, the scalar speed average will reflect the high average speed, and the vector direction average will point out the prevailing wind direction.
A time composite grid cell will contain data if a minimum number of observations exist:
|Averaging Time||Potential # observations||Typical # observations||Minimum # observations|
Data on daily maps are overwritten at both the high latitudes where successive orbits cross and at the "seam" or region where the last orbit of the day overlaps the first orbit of the day.
The center of the first cell of the 1440 column and 720 row map is at 0.125 E longitude and -89.875 latitude. The center of the second cell is 0.375 E longitude, -89.875 latitude.
The data values fall between 0 and 255. Specific values have been reserved:
|0 to 250 =||valid geophysical data|
|251 =||not used for scatterometers|
|252 =||not used for scatterometers|
|253 =||scatterometer observations exist, but are bad|
|254 =||no scatterometer observations|
|255 =||land mass|
The data values between 0 and 250 need to be scaled to obtain meaningful geophysical data. To scale the data:
|TIME:||either multiply by||6.0||to get||0 to 1440 minute of day UTC|
|or multiply by||0.1||to get||0.0 to 24.0 hour of day UTC|
|Wind Speed||multiply by||0.2||to get||0.0 to 50.0 m/sec|
|Wind Direction||multiply by||1.5||to get||0.0 to 360.0 degrees|
|Rain Flag||extract first bit||to get||0 = no rain; 1 = rain|
|Radiometer Rain||extract bits 3 to 8||(x/5)-0.2||to get||0 to 12.5 mm/hr|
|Sum-Of-Squares||multiply by||0.02||to get||a non-dimensional number|
Wind directions are provided in oceanographic convention. This means:
|Winds blowing North:||0° (or 360°)|
|Winds blowing East:||90°|
|Winds blowing South:||180°|
|Winds blowing West:||270°|
The Rain byte contains 3 pieces of information. Use bit extraction to obtain the following:
|scatterometer rain flag
|0 = no rain
1 = rain
|collocated radiometer flag
|0 = no radiometer data within 180 minutes
1 = radiometer data within 180 minutes
|radiometer columnar rain rate
(bits 3 - 8)
|0 = no rain
1 = rain in adjacent cells
2 thru 25 = value/5 -0.2 rain rate in mm/hr
The files are stored in zipped form. If your programming environment does not read compressed files directly, use any GZIP compatible tool to unzip files before reading.
Read routines are available from the FTP directory. Read routines for the orbit (L2B) and gridded data are available in MatLab, IDL, Fortran, C++ and Python. Be sure to read the ReadMe.txt files in the associated directories.
There are gaps within these data. Missing data generally affect Daily and 3-Day products, but can also reduce the number of observations in Weekly and Monthly averages. The following are known missing days in the RSS ASCAT data set.
April 21-24, 2007
September 18, 2007
January 17, 2008
March 20, 2008
May 15, 2011
When browsing imagery, the navigation may skip dates with no data, or you may see a blank map stating that no data are available for that time. Binary data files for dates with completely missing data are not produced; they will be absent from our FTP server.
Data gaps are generally due to missing data upstream from our processing facility, such as the instrument being turned off. Occasionally, there are delays in obtaining and/or processing recently recorded data; beyond several weeks, it is unlikely that missing data will become available.
Official information on missing ASCAT data can be found at the OSI SAF web site.
Each daily, weekly, or monthly ASCAT image in our browse data section shows the wind speed and direction for a specific geographical region.
The daily browse images display the ascending and descending satellite passes separately. The approximate UTC time of each pass is labeled near the bottom of the image. The date of the data displayed is the UTC date when the data was collected (See the Map dates and Times section of the FAQs). The observation times of ascending and descending pass segments are interleaved throughout the day. When browsing daily pass segments with the Previous and Next buttons, approximately half of the Earth will be browsed in temporal order; the other half will not be browsed in temporal order.
A scale of 10 meter ocean surface wind speeds is located on each image and extends between zero and 45 m/s. Land regions are colored gray. Areas where scatterometer data are not available are black. For the daily maps, the black color includes areas where the satellite did not pass over and no data was collected, areas where data was collected but it was determined to be bad, coastal areas, and regions containing sea ice.
The browse images are produced from the same gridded data files available on our FTP server.
Swath Data Files
In addition to the gridded daily and time-averaged files, we release swath files only for scatterometer data. These files contain the measurements for one orbit of the satellite around the Earth. The data are organized by observation cells that are perpendicular to the direction the satellite travels. This makes the data grid at an angle to the standard lat/lon grid used in our other products.
Orbital scatterometer data are publicly available via FTP at: ftp://ftp.remss.com/ascat/MetopA/ascat_wind_vectors_v02.1
The files are stored in directories based on orbit number (00000to09999, 01000to01999, 02000to02999, etc.). The file names have the form: winvec_RRRR_v02.1.dat.gz where RRRRR is the five digit orbit number.
The orbital data file format is described at: ftp://ftp.remss.com/ascat/readme_ascat.txt
Read routines for the orbit (swath) files are provided in Fortran, IDL and Matlab at: ftp://ftp.remss.com/ascat/ascat_orbit_support/
Verification data to help ensure you are reading the data correctly are included in the readme file at: ftp://ftp.remss.com/ascat/ascat_orbit_support/readme.txt
Figa-Saldaña, J., J. J. W. Wilson, E. Attema, R. V. Gelsthorpe, M. R. Drinkwater and A. Stoffelen, (2002) The Advanced Scatterometer (ASCAT) on the Meteorological Operational (MetOp) Platform: A Follow on for European Wind Scatterometers, Canadian Journal of Remote Sensing, 28(3), 404-412.
Ricciardulli, L. and F. J. Wentz (2014), Integrating Multiple Scatterometer Observations into a Climate Data Record of Ocean Vector Winds, paper presented at 2014 Ocean Sciences Meeting, held in Honolulu, Hawaii, USA.
Ricciardulli, L. and F. J. Wentz (2012), Development of Consistent Geophysical Model Functions for Different Scatterometer Missions: Ku and C-band, paper presented at the NASA IOVWST meeting held in Utrecht, Netherlands.
Ricciardulli, L., T. Meissner and F. J. Wentz (2012), Towards a Climate Data Record of Satellite Ocean Vector Winds, in Proceedings of the IEEE International Geoscience and Remote Sensing Symposium, Munich, Germany, pgs 2067 - 2069.
Related Data Sets
Other ASCAT data are available from the European Space Agency Data Center produced by KNMI with a mirror of the data at the Jet Propulsion Lab PO.DAAC. An alternative version is produced in real-time by NOAA/NESDIS/STAR. Distinctly different GMFs, wind algorithms and quality control steps are used in making these ther data sets.
ASCAT data were produced by Remote Sensing Systems and are funded by the NASA Ocean Vector Winds Science Team. We are thank EUMETSAT for providing the ASCAT SZR L1B data files used in making these products.
How To Cite ASCAT Data
Production of this data set could not have happened without support from NASA. We need you to be sure to cite these data when used in your publications so that we can demonstrate the value of this data set to the scientific community. Please include the following statement in the acknowledgement section of your paper:
"C-2015 ASCAT data are produced by Remote Sensing Systems and sponsored by the NASA Ocean Vector Winds Science Team. Data are available at www.remss.com."
An official data citation for use in publications is given below. Insert the appropriate information in brackets as needed.
Ricciardulli, L., Wentz, F.J., April 2016. Remote Sensing Systems ASCAT C-2015 Daily Ocean Vector Winds on 0.25 deg grid, Version 02.1, [indicate subset used]. Santa Rosa, CA: Remote Sensing Systems. Available at www.remss.com. Accessed [dd mmm yyyy].