Known Issues

Inter-Satellite Comparisons of Environmental Parameters

At RSS, we make every effort to carefully intercalibrate and consistently process radiometer and scatterometer data and to provide climate-quality retrievals of environmental parameters throughout the satellite era, despite specific satellite platforms and instruments being placed into and taken out of orbit and operation. The Version-7 products currently available at RSS are our most accurate parameters to date.  That said, there remain instrumental or orbital characteristics that make the agreement between individual instruments difficult to maintain over time.  As a specific instrument ages, the quality of data and information returned to earth may degrade, which can impact our environmental parameter retrievals. Certain degradations can be compensated for, others can not. We do our best to keep track of instrument degradation and artificial drifts in the data time series. One particular method that is helpful for identifying issues with a specific instrument is the intercomparison of satellite data during overlapping periods of operation.

TMI was recently turned off (April 8th, 2015) and we just released our fully reprocessed TMI data set (more than 17 years of data).  We brought the TMI data up to the RSS Version-7 data processing standard. During the re-calibration stage, we completed a careful intercomparison of TMI data to that from all RSS MW radiometers .  The findings of our intercomparison between TMI retrievals of SST, Wind Speed, Water Vapor, Cloud Liquid Water, and Rain Rate with those from SSM/I, SSMIS, AMSR-E, AMSR2, and WindSat instruments are shown below.  Our comparisons show no problems with TMI over the entire operational life cycle with the exception of the last 9 months of operation during which time the altitude of the satellite was rapidly changing.  The analyses consist of time series of monthly differences over available instrument overlap periods and show the differences of the other MW sensor data minus the TMI data.  A collocation window of ±1 hour and ±25 km is used.  For all MW sensors, we use the same retrieval algorithm, Version-7, to process the brightness temperatures to environmental parameters.  This helps to achieve consistency among the different sensors.  The retrieval algorithm is described by Wentz and Spencer (1998), Chelton and Wentz (2005), and Wentz and Meissner (2007).  The rain rate retrieval part of the algorithm is further detailed in Wentz and Spencer (1998) and Hilburn and Wentz (2008a). 

Figures 1 through 5 show the time series of the monthly averages of the differences for SST, Wind Speed, Water Vapor, Cloud Liquid Water, and Rain Rate, respectively.  For each time series, these figures show the monthly mean values of the difference, called offset, and the drift in the difference which is the least-squares slope of the time series times the duration of the time series and hence is a measure of the change in parameter over the period of overlap.  For Figure 1, which shows SST, there are only 3 other sensors that provide SST retrievals: WindSat, AMSR-E, and AMSR2.  For the other figures, there are a total of 9 MW radiometers that are compared to TMI.  Furthermore for Figure 2, Wind Speed, there are two additional sensors: the QuikSCAT and ASCAT scatterometers.

Given this many comparisons, certain problems with some of the MW radiometers become obvious.  These problems are summarized in Table 1 shown below.  The most notable problems are with the F15 SSM/I and the F16 SSMIS instruments. 

On August 14, 2006, a radar calibration beacon (RADCAL) was activated on the F15 satellite platform.  Although we have taken measures to correct this problem (Hilburn and Wentz 2008b; Hilburn 2009), the retrievals are still quite noisy during the RADCAL period (shown in red in Figures 2 through 5). 

Starting at 2009, the F16 SSMIS retrievals begin to significantly degrade.  The F16 SSMIS is a problematic sensor as it has an emissive antenna, sun intrusion into the hot load, and an orbit with a rapidly drifting ascending node time.  The calibration of F16 needs to be revisited.  

Of minor note is that the last year of F14 SSM/I operation (2008), there are anomalous retrievals as shown in red in Figures 2 through 5.  We do not know the cause of these degraded retrievals. 

In all cases, the portions of the time series that are marked in red are excluded from the computation the offset and drift for that instrument.  The offset and drift is listed at the top of each plot.

We do not find any problems with TMI, which seems to be an extremely stable sensor.  However, since we are looking at monthly averages in this analysis, problems associated with the TMI emissive antenna likely average out and are not noticeable.  The comparisons of SST (Figure 1) show excellent agreement as do the Wind Speed comparisons (Figure 2).  The Wind Speed time series plots also include the QuikSCAT and ASCAT scatterometers, which measure wind speed in an independent manner from radiometers and are a good check on the TMI winds.  Apart from the issues listed in Table 1, the Water Vapor, Cloud Liquid Water, and Rain Rate comparisons (Figure 3, 4, 5) also look very good for TMI.  There is no suggestion of a discontinuity between the pre-boost TMI retrievals and the post-boost retrievals (the boost occurred August 2001).  In Figure 2, the very small increase in the Wind Speed time series during 2014 for both WindSat and AMSR2 as compared to TMI is likely due to the rapidly decaying TRMM orbit.  The TRMM started descending towards Earth in mid-2014, and the change in altitude from 405 km to 350 km spuriously decreased the TMI wind retrievals by 0.05 to 0.1 m/s.  Surprisingly, the comparison with ASCAT does not show this difference.

Results presented here have been recently submitted for publication to the Journal of Climate.

Table 1.  Summary of Findings from Environmental Parameter Comparisons for RSS Version-7 MW Radiometers to TMI

Sensor

Known Issue

TMI

Comparisons do not imply any problems with TMI, with the exception of Wind Speed after August 2014 (last 6 months of operation)

F11 SSM/I

Small offsets in Wind Speed, Water Vapor, Cloud, Rain Rate

F13 SSM/I

Drift of -0.11 m/s/overlap period in Wind Speed; small offsets in Water Vapor, Cloud Liquid Water, Rain Rate

F14 SSM/I

Drift of -0.12 m/s/overlap period in Wind Speed; small offset in Rain Rate; problems during last year of mission (2008)

F15 SSM/I

Noisy retrievals after July 2006 due to RADCAL beacon operation

F16 SSMIS

Problems with Wind Speed, Water Vapor, Cloud Liquid Water after 2009

F17 SSMIS

Drift of -0.2 m/s in Wind Speed, 0.14 mm in Water Vapor, and 0.002 mm in Cloud Liquid Water

WindSat

Offset of 0.006 mm/hr in Rain Rate

AMSR-E

A transition in all parameters near the beginning of 2008

AMSR2

Offset of 0.009 mm/hr in Rain Rate

QuikSCAT

No Wind Speed differences

ASCAT

No Wind Speed differences
 

 

Fig. 1. Time Series of Monthly SST Retrievals from MW Radiometers Compared to TMI

 

 

Fig. 2. Time Series of Monthly Wind Speed Retrievals from MW Radiometers Compared to TMI.  Red regions of the time series represent problem periods for that sensor.

 

 

Fig. 3. Time Series of Monthly Water Vapor Retrievals for MW Radiometers Compared to TMI.  Red regions of the time series represent problem periods for that sensor.

 

Fig. 4.  Time Series of Monthly Cloud Liquid Water Retrievals for MW Radiometers Compared to TMI.  Red regions of the time series represent problem periods for that sensor.

 

Fig. 5. Time Series of Monthly Rain Rate Retrievals for MW Radiometers Compared to TMI.  Red regions of the time series represent problem periods for that sensor.

References