Benefits of interpreted vector programming and Hierarchical Data Format for statistic ocean model evaluation

Abstract

HPC world is increasing incredibly fast.We can build more resolute and accurate models. Unfortunately,Big Data management pops up when trying to evaluate them.We set up a Python module that does automated assessment of a set of observations,extraction and aggregation in time of ocean model time,running evaluation methods. For the first time execution,we used ocean physics variables and CMEMS reprocessed moorings historical data,two model data from the RITMARE and from CMEMS. The results are promising,because we can process an historical or real time evaluation and it is portable,the evaluation results provides more reliable skill scores.