Investigating Short-Wavelength Correlated Errors on Low-Resolution Mode Altimetry

The observation of ocean scales smaller than 100 km with low-resolution mode (LRM) altimetry products is degraded by the existence of a hump artifact visible on sea surface height (SSH) spectra. Through an analysis of simulations and actual data from multiple missions, this paper shows that the hump originates in a response to inhomogeneities in backscatter strength. Current retrackers cannot fit their Brown model properly because they were designed for a scene with homogeneous backscatter properties. The error is also smoothed along track because of the size and shape of the LRM disc-shaped footprint. Therefore, the hump is modulated by the altimeter design and altitude and by the retracker used. Because of the random nature of the phenomenon, a large majority of long topography segnients (e.g., hundreds to thousands of kilometers) is affected. However, within these segments, a substantial fraction of the corruption is contained in small subsets of data (e.g., less than 10%). This paper shows that oceanography users interested in small-scale SSH signals can mitigate the hump corruption by using better editing and postprocessing algorithms on the 20-Hz rate of current products. Last, the thin stripe-shaped footprint of Cryosat-2's synthetic aperture radar mode (SARM) is not affected by the hump artifact, thus improving the observation of topography features ranging from 30 to 100 km. The differences between SARM and pseudo-LRM sigma0 can also be used to detect major hump events on pseudo-LRM data, which might be an asset to design/validate a new generation of algorithms aimed at reducing the hump artifact on the existing LRM record.