Mohit Anand

Mohit Anand

Doctoral Candidate

Address


Permoserstraße 15
04318 Leipzig

Building 7.1, Room 414

mohit.anand@ufz.de


About Me


I work on identifying compound climate drivers of high impact events using deep learning. Currently I am working on finding climate drivers of extremely high forest mortality. I train deep learning models like Variational Auto-Encoders , Transformers, Diffusion Models, etc and try to modify their structure to make them more eXplainable. I started my PhD in Compound weather and climate events group under the supervision of Prof. Dr Jakob Zscheischler.

Education


  • 2020-Present: Doctoral Candidate
  • 2017-2020: MSc in Environmental Engineering, ETH Zurich, Switzerland
  • 2013-2017: B.Tech in Environmental Engineering, IIT(ISM) Dhanbad, India

Projects


  • Identifying weather drivers of extremely high forest mortality given forest structures like age, diameter, leaf area index of trees using Transformers
  • Adding diffusion models to the benchmark on WeatherBench dataset
  • Identifying the role of winter climate on extreme summer leaf area index over entire northern hemisphere using structural equation modelling
  • Generating unseen weather prototypes of extremely high forest mortality using Variational Auto Encoders
More information