Brief Description of the OTED

A triple collocated dataset is created from the OceanSat Scatterometer-2 (OSCAT) normalized radar backscatter (sigma0), the Tropical Rain Measuring Mission (TRMM) near-surface rain rate, and the European Centre for Medium-Range Weather Forecasts (ECMWF) near-surface vector winds (VW). The dataset is created using measurements from the 3 data sources that are temporally and
spatially within 30 minutes and 1 km respectively. The dataset spans the entire mission of OSCAT available OSCAT data (2011 to 2014). 

OSCAT operated between 2009 and 2014 [1]. It used a Ku-band rotating pencil beam radar to measure the surface (sigma0). There are two beams of measurements, an inner beam and an outer beam, measured at two incidence angles, 49^o and 58^o. OSCAT has a polar orbit which covers the entire earth. Three OSCAT data types are used: Level-2B (L2B), Level-1B (L1B) and Ultra High Resolution (UHR). The L1B contain the raw sigma0 data, while the L2B contain the VW estimates, and the UHR contain high resolution data. The L2B data are spatially sparser than the L1B making them faster to parse than the L1B data. The UHR estimates high resolution sigma0 values on a 2.5×2.5 km grid using the antenna response function. TRMM operated on a Japanese satellite between 1997 and 2015 [2]. Similar to OSCAT, TRMM also operates a Ku-band radar but measures the response to rain. The incidence angle of TRMM ranges between -18 to 18 degrees. TRMM has a precipitation radar that measured the near-surface rain rates. The coverage of TRMM is restricted to near the equator. ECMWF uses numerical weather prediction (NWP) to generate near-surface vector winds [3]. ECMWF hasa product that estimates the near-surface global wind for times a day (Midnight, 6 AM, Noon, and 6 pm) on 50 km grids. The ECMWF VW are interpolated in space and time for specific locations in the collocation process.

The method for collocating the datasets consists of two steps. First is an initial coarse collocation between TRMM and OSCAT L2B data locate where the TRMM and OSCAT orbits are within 30 minutes and 1 km. Around the coarse collocations a grid of UHR simga0 values are produced. Second, a fine collocation is computed between the OSCAT (L1B, L2B, and UHR) and TRMM data. The same temporal and spatial requirements from the coarse collocations are used for the fine collocations. The ECMWF VW data is computed using trilateral interpolation for the times and locations of the finely collocated OSCAT and TRMM data.

The OTED is divided into 6 arrays A - F . Array A contains the primary scientific data needed to create the SWR GMF. The contents of A include the OSCAT sigma0 for both the L1B and UHR. Along with the sigma0, the A also includes the incidence angle of the OSCAT L1B and UHR1. The azimuth angle is defined as being the clockwise angle between true north and the direction the antenna is pointing [4]. The azimuth angle is included (which should be distinguished from the antenna azimuth angle, which is actually included in D) for both L1B and UHR. The time and location of each specific L1B and UHR measurement is included in B. The non-OSCAT data included in A is the TRMM surface R and PIA. The TRMM PIA measures the attenuation of the signal as it passes through the atmosphere. The ECMWF wind speed and wind direction are calculated from the ECMWF U and V components. The time, latitude and longitude are the TRMM geographical location and times.

The five other arrays (B, C, D, E, and F ) of the OTED contain supplemental information from the collocations. This data is not necessary for the SWR GMF creation but is useful for SWR GMF validation and in other atmospheric studies. The B array contains data about the temporal and spatial differences between L1B measurements and TRMM. The C array contains data about the time and location of the OSCAT measurements. The D array contains auxiliary OSCAT data like relative azimuth angle and normalized standard deviation of the communication noise of sigma0 measurements (Kpc). The E array contains supplemental VW data such as, ISRO L2B VW estimates, and U and V components from ECMWF. Finally, the F array contains the WO estimates from OSCAT sigma0 and theoretical WO sigma0 calculated from ECMWF VW.


The data in the columns of each array is described in the longer document. 

[1] SCAT-DP TEAM, “Oceansat-2 Scatterometer algorithms for sigma-0, processing and products format,” user manual, Advanced Image Processing Group, Signal and Image Processing Area, Space Applications Centre, Gujarat, India, 4 2010.

[2] Precipitation Radar Team, “Tropical Rainfall Measuring Mission (TRMM) Precipitation Radar Algorithm,” user manual, TRMM Precipitation Radar Team (JAXA and NASA), USA, 7 2011. 

[3] A. Persson and F. Grazzini, “User guide to ECMWF forecast products,” Meteorological Bulletin, vol. 3, no. 2, 2007. 

[4] National Oceanic and Atmospheric Administration (NOAA), “WindSat Data Processing Documentation,” Technical Report WindsatDocJan06-1, National Oceanic and Atmospheric Administration (NOAA), 2006. Accessed on 2024-05-07.