Deep convection occurs in oceanic regions submitted to strong atmospheric buoyancy losses and results in the formation of deep water masses (DWF) of the ocean circulation. It shows a strong interannual variability, and could drastically weaken under the influence of climate change. In this study, a method is proposed to monitor quantitatively deep convection using multisensors altimetry and ocean color satellite data. It is applied and evaluated for the well-observed Northwestern Mediterranean Sea (NWMS) case study. For that, a coupled hydrodynamical-biogeochemical numerical simulation is used to examine the signature of DWF on sea level anomaly (SLA) and surface chlorophyll concentration. Statistically significant correlations between DWF annual indicators and the areas of low surface chlorophyll concentration and low SLA in winter are obtained, and linear relationships between those indicators and areas are established. These relationships are applied to areas of low SLA and low chlorophyll concentration computed, respectively, from a 27 year altimetry data set and a 19 year ocean color data set. The first long time series (covering the last 2 decades) of DWF indicators obtained for the NWMS from satellite observations are produced. Model biases and smoothing effect induced by the low resolution of gridded altimetry data are partly taken into account by using corrective methods. Comparison with winter atmospheric heat flux and previous modeled and observed estimates of DWF indicators suggests that those DWF indicators time series capture realistically DWF interannual variability in the NWMS. The advantages as well as the weaknesses and uncertainties of the method are finally discussed.