News

Presentations given in “Big data and machine learning in geoscience” session at EGU 2019

Bing Gong gave her presentation on “Prediction of daily maximum ozone threshold exceedances by artificial intelligence techniques in Germany” and Felix Kleinert presented “Near Surface Ozone Predictions Based on Multiple Artificial Neural Network Architectures” in the session on Big data and machine learning in geosciences at the European Geosciences Union General Assembly in Vienna, Austria.

Parallel workflow for data extraction developed

Martin Schultz and Severin Hußmann implemented a parallel workflow for extracting arbitrary fields from large numerical weather model data files for use in machine learning applications. This workflow uses the JUBE tool developed at JSC for benchmarking purposes.

Presentation of results on using SMOTE for ozone prediction

At the Computational and Data Science (CaDS) Seminar, JSC, on Tuesday, 19th Mar 2019, Bing Gong gave a presentation about the preliminary results of her latest research on using the synthetic minority oversampling technique (SMOTE) to improve ozone threshold exceedances prediction with various machine learning techniques.

Workshop: AI for weather and climate studies

Bing Gong attended the Eumetnet workshop on “AI for weather and climate studies” held at Royal Meteorological Institute of Belgium. In the two-day workshop, 20 presentations were given on the existing or exploring potential weather and climate application of AI. Bing Gong gave a presentation on “Parallel deep learning workflow for short-term temperature forecasts with video frame prediction methods”.

Presentation accepted at EGU 2019

Abstract “Near Surface Ozone Predictions Based on Multiple Artificial Neural Network Architectures” by Felix Kleinert is accepted at EGU as oral presentation on Monday, 8th of April 2019.

High-resolution topographic data imported into Rasdaman

Lukas Leufen and Jessica Ahring successfully imported sample high-resolution topographic data from the 90 m TanDEM-X data product into the Rasdaman data base at JSC, which will be used to store and retrieve geophysical data for the IntelliAQ project.

auto-qc to be developed for quality control of ozone time series data

A novel concept for automated quality control of the global air quality data from the Tropospheric Ozone Assessment Report (TOAR) database has been proposed. The approach seeks to quantify the quality of individual measurements after applying a set of statistical tests that are in use in several environmental agencies such as EPA. To prove the concept, we have started to develop a software package (auto-qc) for double-checking the quality of the multi-annual hourly ozone time series from the TOAR database.