Day 4

Detailed paper information

Back to list

Paper title Factors Affecting Altimeter Sigma0
Authors
  1. Graham Quartly Plymouth Marine Laboratory Speaker
Form of presentation Poster
Topics
  • A8. Ocean
    • A8.13 Remote-sensing of Ocean Winds and Stress
Abstract text Over the ocean, an altimeter's measurements of normalised backscatter (sigma0) are interpreted as a measure of surface roughness, which is linked empirically with wind speed. However there are many other factors, both instrument specific and environmental, that affect the observed values. Observations at different radar wavelengths record the sea surface roughness at different spatial scales; I utilise the close relationship between backscatter strengths at the two most common altimeter frequencies (Ku-band and C-band) to highlight that other factors are present. Because of their differing sensitivities to scales of roughness, the sigma0 difference, sigma0_0Ku miinus sigma0_C, is not a simple offset, but has a peak for wind conditions around 6m/s. Instrumenting and processing options tender to alter the sigma0 values uniformly across a wide range of values, whereas environmental conditions tend to affect the shape of the relationship, as they are altering the interplay of different roughness scales or their radar-scattering properties.

As there is no universally recognised absolute calibration of altimeter sigma0 values, all instruments require a simple bias to bring them to a common scale. However there is also a dependence on the retracking algorithm used, with the Maximum Likelihood Estimator, MLE3, giving fairly robust estimates, whereas MLE4 and the standard SAMOSA retracker for SAR waveforms show a strong dependence on the inferred mispointing, and so need an adjustment correction for that. It has also been noted that the TOPEX and Jason altimeters in their non-sunsynchronous orbit have additional biases with a period of 58.77 days linked to the degree of solar exposure of the instrument in orbit.

There is an atmospheric effect that changes the sigma0_Ku values, and that is attenuation by liquid rain. This is only significant for a small percentage of observations, but the effect is more pronounced at the Ka-band of AltiKa and the future CRISTAL mission. The most marked environmental effect is due to wave height. At very low winds, a change in wave height of 1m can affect sigma0 by ~0.15 dB, but this causes minimal bias in wind speed estimates due to the low sensitivity to sigma0 of all wind speed algorithms in this regime. There is also an effect at high wind speeds which remains to be accurately quantified. Finally the sigma0_Ku-sigma0_C relationship appears to shift by about 0.15 dB on moving from tropical to polar waters. This confirms a previously reported temperature effect, although the size of the change is a little less than theoretically expected.

All these factors are reviewed within the scope of efforts to remove biases in wind speed algorithms that are either regional in nature or vary with satellite, so as to further efforts to develop a homogeneous altimetric wind speed product.


Figure shows the mean sigma0_Ku-sigma0_C relationships for 4 current altimetric satellites. On the left are the curves for each after initial bias adjustments, showing a divergence in behaviour at very high sigma0 (low winds); on the right are the variations of each curve with sea surface temperature.