Introduction
RapidScat was installed on the Columbus module of the International Space Station (ISS) in September 2014 and has been making microwave observations of the Earth surface since the beginning of October 2014. This radar instrument operates in a very similar manner to QuikSCAT, a previous Ku-band scatterometer. More information is available at the mission website. RapidScat wind data are publicly available from the PO.DAAC. Some of the plots shown below use the PO.DAAC winds, and some use an RSS version that we are preparing for distribution.
Current RapidScat Studies at Remote Sensing Systems
Geophysical Model Function Development
RSS has developed Geophysical Model Functions (GMFs) for satellite microwave scatterometers since the early 1980's. The GMF currently used to process the science-quality RapidScat microwave backscatter measurements into ocean wind speeds and directions was developed by RSS and is an extension and improvement of the QuikSCAT Ku-2011 GMF (Ricciardulli and Wentz, 2011; Ricciardulli and Wentz, 2015). We expect this RapidScat GMF to be refined in the coming months as more RapidScat data become available.
Production of Climate-Quality Winds
RSS specializes in processing satellite microwave data and focuses on constructing climate-quality ocean wind data sets. Our goal is to provide users with scatterometer and radiometer winds that are inter-calibrated and consistently processed so that they can be combined into an ocean wind speed product suitable for use in long-term study. A presentation given at the American Geophysical Union Fall Meeting in 2014 (Ricciardulli, Meissner and Wentz, 2014; 2012) summarizes our work on the development of the Climate Data Record of ocean winds. Ocean vector winds consist of wind speeds and directions and are only available from scatterometers and the one polarimetric radiometer in operation, WindSat. Currently, we offer via our web site and ftp server, QuikSCAT, WindSat and ASCAT ocean vector winds processed by RSS. RapidScat winds will be added sometime in mid-2015, as well as a new recalibrated version of ASCAT winds. In addition to the ocean vector winds, we have radiometer wind speeds available from 1987 to the present available.
Wind Validation
RSS has been a significant partner in the RapidScat calibration and validation team directed by Ernesto Rodriguez at JPL. During the immediate post-launch period, our analyses and comparisons of RapidScat winds with other RSS-processed satellite winds assisted in the intercalibration of the RapidScat data. As an example of the quality of the PO.DAAC RapidScat winds, the following images show global maps of the average wind speed differences (top) and the standard deviations between RapidScat and WindSat rain-free wind speeds (bottom) using a 90-minute colocation window and the first 6 months of data (Oct 2014 - Mar 2015). For much of the global ocean, RapidScat and WindSat winds are in agreement and have a standard deviation less than 1 m/s. Regions of high winds are more susceptible to small biases and greater uncertainty.
Comparisons of RapidScat versus other RSS-processed satellite winds help to ensure the stability of the RapidScat timeseries. The following plots show comparisons of PO.DAAC RapidScat winds versus WindSat (top) and AMSR2 (bottom) wind speeds (90 min, rain-free colocations). These radiometer winds are in excellent agreement with RapidScat over the short period of operation (6 months). There are no obvious trends or great differences between the wind speeds despite the differences in measurement styles and data processing of the scatterometer and radiometers. The average wind speed varies over time due to the changes in RapidScat non-sun-synchronous measurement collocations with other sun-synchronous satellites. The orbit of the ISS causes RapidScat to view differing scenes daily, with a one month variability between higher latitudes (and higher wind speeds) and lower more tropical latitudes (with lower mean wind speeds).
Construction of Ancillary Data Swath Files
To assist with validation and rain correction development, we have created ancillary swath data files that contain RapidScat winds tightly colocated with wind and rain observations from the various RSS V7 radiometers (SSM/I, SSMIS, AMSR2, and WindSat) currently in operation. We use a maximum time window of 3 hours. Radiometer winds are valid only in rain-free areas, except for WindSat which can retrieve wind speeds in all weather conditions, even storms, due to a special all-weather algorithm and the 6 GHz channel (not available on other radiometers). The ancillary files therefore contain both wind and rain radiometer products. When rain is present, there will likely be no radiometer wind value if WindSat is not within the colocation window. When there is no rain, the radiometer winds are included.
One ancillary data file is provided for each RapidScat orbit and includes:
- the closest-in-time radiometer wind speed observation
- the closest-in-time radiometer rain observation
- the time difference between RapidScat and the radiometer observation
- an index value that identifies which radiometer was the closest in time
- latitude and longitude of the measurements
- the RapidScat wind speed
A pdf of Ancillary Data File contents contains plots for orbit 628 as an example.
We delivered the ancillary swath data files to JPL for use in the RapidScat science data development phase. The RapidScat group at JPL used these ancillary data to develop a rain correction to the RapidScat 12 km data product, which is distributed through the PO.DAAC.
These files can also be used for wind validation and are available to scientists upon request.
Stability of Winds
The consistently and carefully processed RSS wind products are intercompared over time in order to monitor the stability of each instrument. RapidScat winds are the latest products added to our microwave wind data series. The figure below shows the consistency of the RSS climate-quality scatterometer and WindSat winds, in terms of the stability of the wind speed differences over time between the RSS recalibrated ASCAT winds versus collocated QuikSCAT winds, RSS-processed RapidScat winds, and RSS WindSat (4 hr colocations) winds. The plot is for the period 2007-2015 and consists of data averaged over the global oceans from 55S to 55N. For climate analysis, wind speed differences on monthly global scales as shown here must be no more than 0.1 m/s. The plot shows that this is the case over all instrument lifetimes. RapidScat has only been operating for a few months and we expect that future reprocessing will further improve the agreement between it and the other wind vector data.
Diurnal Studies
The sampling of RapidScat at all times of day allows for the study of the diurnal variation of ocean winds. We have previously looked at wind speeds throughout the day using TMI, but this is the first time we can explore the changes in wind speed and direction using a scatterometer. Preliminary analyses of the diurnal cycle are underway.
Related Data Sets
The RapidScat L2B (swath) data are available from the JPL Physical Oceanography DAAC (PO.DAAC).
References
- Ricciardulli, L. and F.J. Wentz, 2015: A Scatterometer Geophysical Model Function for Climate-Quality Winds: QuikSCAT Ku-2011. Submitted in 2014 to Journal of Atmospheric and Oceanic Technology.
- Ricciardulli, L., T. Meissner, and F. Wentz, 2014: Building a Climate Data Record for Ocean Vector Winds, abstract OS51D-02, invited talk presented at 2014 Fall Meeting, AGU, San Francisco, CA, 15-19 Dec.
- Ricciardulli, L., T. Meissner, and F. Wentz, 2012: Towards a climate data record of satellite ocean vector winds, in Proceedings of the 2012 IEEE International Geoscience and Remote Sensing Symposium, 2067-2069.
- Ricciardulli, L., and F. Wentz, 2011: Reprocessed QuikSCAT (V04) Wind Vectors with Ku-2011 Geophysical Model Function, Technical Report 043011, Remote Sensing Systems, Santa Rosa, CA, p. 8.
Acknowledgements
RapidScat GMF development and research is completed by Remote Sensing Systems thanks to funding from the NASA Ocean Vector Winds Science Team and the RapidScat Science and Cal/Val Team (JPL). We are grateful to NASA Jet Propulsion Laboratory (JPL) for providing access to RapidScat L2A and L2B data files.