In-Situ Processing in Climate Science

With climate simulations and earth observations, earth system sciences belong to the most data intensive scientific disciplines, and the rate at which the data is produced increases continuously. Current models supporting a higher complexity paired with an increased resolution produce more and more data that needs to be analyzed and understood. The development of alternatives to the classic post processing/visualization pipeline are therefore mandatory and discussed within this paper, with a strong focus on in-situ visualization and in-situ data processing. Although the work described here is work in progress, large parts are already implemented and tested and on the verge to be deployed in production mode.