Classical water stable isotopes (dD and d18O) have been used for more than 50 years with the aim to understand the links between water cycle and climate. They provide information on either temperature or precipitation changes depending on the latitudes. Their combination, in the so-called d-excess, brings some information on climatic conditions occurring during non equilibrium processes along air masses histories (evaporation over the Oceans, reevaporation of droplets in convective systems, continental recycling or ice crystals formation). Recently, the possibility to measure with high precision d17O in water has enabled to introduce a new parameter, 17Oexcess, resulting from the combination of d18O and d17O. According to both observations and modeling works, this new isotopic parameter is able to decipher some of the non equilibrium processes: when measured in ice core, it is expected to be a more direct tracer of relative humidity of the oceanic evaporative regions than d-excess. In order to better understand what controls this new parameter as well as to extract the maximum climatic information from the combination of 17Oexcess and d-excess, we present different original studies combining these two parameters in several key regions. First, data collected in Niger, West Africa, at scales ranging from the convective system to the seasonal cycle confirm the strong influence of relative humidity on 17Oexcess through the rain reevaporation process. Second, seasonal cycles in the Zongo Valley (Tropical Bolivia) suggest that rain recycling along air masses trajectories have different signatures on d-excess and 17Oexcess leading to decipher the different processes. Third, we study how local processes (precipitation, sublimation) in polar region (Greenland) can affect 17Oexcess archived in ice core with respect to d-excess records through (1) isotopic measurements of vapor versus precipitation collected at the NEEM station and (2) seasonal cycles measured from snow pits.