Assets for SMUrF v1

The SMUrF model [Wu et al., 2021, GMD] leverages satellite-based Solar-Induced Fluorescence data along with an artificial neural network model to estimate biogenic CO2 fluxes over urban areas around the globe.


The model code with the latest commit is available on GitHub.


Oak Ridge National Lab DAAC is performing the quality check and will soon release the SMUrF flux output. If you are interested in the early access, feel free to contact Dien. (last update: Jun 14, 2021)

As running the model for the entire globe incurs significant computational cost, we only shared the flux results for the following eight regions:

  • western CONUS (125–95 W, 25–50 N)
  • eastern CONUS (95–65 W, 25–50 N)
  • western Europe (11 W–20 E, 35–60 N)
  • eastern China (100–125 E, 20–50 N)
  • eastern Asia (125–150 E, 30–55 N)
  • eastern Australia (130–155 E, 40–10 S)
  • south America (65–40 W, 40–10 S)
  • central Africa (10–20 E, 10–15 N)

Flux components include:
(1) 4-day mean CSIF-based GPP and uncertainty,
(2) daily mean NN-based Reco, and
(3) the final hourly GPP, Reco, and NEE fluxes.


Wu, D., Lin, J. C., Duarte, H. F., Yadav, V., Parazoo, N. C., Oda, T., and Kort, E. A.: A model for urban biogenic CO2 fluxes: Solar-Induced Fluorescence for Modeling Urban biogenic Fluxes (SMUrF v1), Geosci.Model Dev., 14, 3633–3661,, 2021.