Logo Universitat Bremen

Dr. Nasrollah Moradi





+49 421 65788



MARUM, Pavillon 0040

Nasrollah Moradi


I am a computational biophysicist working as a modeller & data scientist in the field of marine science. With the help of mathematical modelling, computer simulation, machine learning and image analysis I leverage and combine field data, laboratory measurements as well as satellite observations to link and better understand small and large scale biogeophysical processes that impact carbon export to the deep ocean

Spe­ci­fic re­se­arch in­te­rests

  • Particle flux and carbon export to the deep-sea
  • Small-scale mass transfer within (and around) phytoplankton colonies and sinking marine aggregates
  • Image analysis and classification of marine particles and zooplankton using machine learning 
  • Computational fluid dynamics and water models 

Current project

Role of Eddies in the Carbon Pump of Eastern Boundary Upwelling Systems (REEBUS): Observation of carbon fluxes in the water column

More details here

Work in progress

A new model for estimating particulate organic carbon (POC) and nitrogen (PON) fluxes vs depth in the water column using measured POC (and PON) from drifting sediment trap samples, in-situ measured particle sinking velocities and image analysis of particles detected by the under water camera profiler (UVP). Modelled POC flux vs measurements (left and middle panels). Modelled POC flux in a mesoscale cyclonic eddy off Cape Verde islands (right panel) 

modelling particulate organic carbon in the water column

Reconstruction of cloudy satellite chlorophyll maps using machine learning as a necessary step in order to correlate ocean surface chlorophyll concentration with carbon export in the deep ocean

Reconstruction of cloudy satellite maps

Development of a custom Python package (Gel-trap Particle Image Size Analyzer) for objective image analysis of marine particles collected using drifting sediment trap: stacking, stitching, background removal and particle extraction & classification 

Gel-trap particle image analysis