Upper Ocean Stratification from Filters for SMOS/Aquarius Cal-Val
[16-Apr-2013] Reverdin, G., Reverdin, G., Boutin, J., Martin, N., Kolodziejczyk, N., Gaillard, F., Rolland, J., and Blouch, P.
Presented at the 2013 SMOS-Aquarius Science Workshop
Since 2005, we have deployed SVP-BS drifters and attached floats to measure ocean temperature and salinity at different depths within the top 60 cm layer in order to provide information for calibrating and validating new L-band radiometer satellite data. Temperature and haline stratification have been investigated from these data both during large daily cycles and also during rain events.
We have collected in 2009-2012 more than 15 years of multi-parameter drifter data that have been validated and that are used to qualify the near-surface stratifications. Close to half these data are in the wet tropics. We find that isolated rainfall event in the tropics can be followed for a few hours by the drifters and that they correspond usually both to temperature and haline stratification in the top 50 cm for at least one hour following the event (between 15 and 50 cm, the difference in salinity averages 20% of the total salinity drop, and in temperature 0.22°C for a 1 psu salinity drop). We confirm from collocated SSMI wind data that stratification and recovery are dependent on wind intensity during and after rainfall, and of course tend to be proportional to the total rainfall rate. However, we are limited in that the wind and rainfall estimates are no simultaneous to the freshwater events.
The multi-sensor drifters also provide information on the temperature stratification in the upper 60 cm top layer that happens for the large daily cycles (typically, for 4% of the days in the tropics and subtropics). For these daily cycles larger than 1°C near the surface, we find that there is a tendency for a delayed cycle at 50 cm or deeper, with a vertical gradient in the mid-day to early afternoon that tends to be nearly homogeneous in this layer, and a strong reduction afterwards near the surface, as the mixing penetrates deeper. For the very large daily cycles, the stratification on the other hand is not linear through this layer. We have also indication that these are associated with a daily cycles in surface salinity with a late afternoon maximum, that even for some relatively small/intermediate daily cycles in T (0.5°C) can reach 0.01 psu. What we used did not allow to estimate the salinity daily cycles for large near-surface vertical gradients.
We recently developed a new small wave rider Surpact that can both measure T and S close to the surface (at 4-5 cm) and short wave spectra. From the spectra, we estimate a wind in the previous hour. This is validated during deployments in the Spurs/Strasse cruise in August/September 2012. One Surpact prototype was then deployed which experiences five rain events. The data are promising and for these cases in the autumn 2012 subtropical North Atlantic there is no significant wind increase during the rainfall events.