Long-term quality and stability assessment of CryoSat-2 Ocean Data

Abstract

Since 8 April 2010 ESA's Earth Explorer mission CryoSat-2 has been measuring marine and land ice thickness variations with a dedicated altimeter. With the effects of a fast-changing climate becoming apparent, it is increasingly important to understand exactly how Earth's ice fields are responding and regulate climate and sea level. It is therefore of utmost importance that the quality of the CryoSat-2 altimeter data meets the highest accuracy and precision level, not only over the ice caps and sea-ice surface but over the oceans as well. The SIRAL altimeter of CryoSat-2 is capable of measuring high-resolution geophysical parameters from the open ocean to the coast. In this paper, we validate the CryoSat-2 geophysical ocean product, Baseline-B (GOP). The analyses focus on long-term monitoring, validation, and cross-calibration of these ocean altimeter data with all the concurrent altimeter data in the Radar Altimeter Database System RADS and with independent in-situ tide gauge measurements from the Permanent Service for Mean Sea Level PSMSL. The objective is to evaluate the stability of the CryoSat-2 measurement system and the identification and explanation of biases and bias drifts. A persistent monitoring of these biases in time is important to establish the data's usefulness for investigation of long-term sea level and ice topography/volume changes. The results constitute the final status of the GOP baseline-B product for use as essential climate variables. GOP ocean data has a -6.3±0.2 cm range bias w.r.t. to all calibrated satellites in RADS and a drift of 0.14 mm yr⁻¹ w.r.t. Jason-2. GOP data has significant wave height and sea state biases close to that of the Jasons’, and seems to be better handling the LRM-SAR transitions than the RADS CryoSat-2 product. GOP data w.r.t. PSMSL tide gauge data has an average correlation of R = 0.83, a mean st. dev. of σ = 6.30 cm, and a drift of 0.11 mm yr⁻¹, equivalent to 40 Gt yr⁻¹ ice sheet mass loss. It is concluded that GOP CryoSat-2 performs better than RADS CryoSat-2, which seems to suffer from a 480-day platform temperature cycle. GOP is also on a par with the altimeter data from the Jason reference missions, and well suited for ocean and climate studies.