A regional-scale modelling system is currently being developed for simulating tranport of greenhouse gases in Northern Alberta. The system will directly relate both ground- and satellite-based observations with regional surface fluxes, using both forward and inverse modelling. One of the goals of inverse modelling is to evaluate the feasibility of the system to estimate major anthropogenic point sources of CO2 associated with heavy industrial development in Northern Alberta.

What you can see here are the first results of test simulations. This is work in progress so don't expect some convincing results or high quality presentation. Some things which are not working well at this stage are pointed out below, some issues (e.g. proper emission modelling) are not addressed in this particular case. Purpose of this simulation is to make sure that everything works as intended and learn a little more about the behaviour of the model.


Current simulation grid is 100x100 cells with 10km resolution. It covers north part of Alberta province in Canada with centre close to Alberta Oilsands industrial area. Domain also includes parts Saskatchewan, British Columbia and Northwest Territories.

Simulation domain with anthropogenic emissions

Simulation domain showing anthropogenic emissions from EDGAR.

Black line is used for province borders. Lake Athabasca can be seen at the north part of border between Alberta and Saskatchewan and a part of Great Slave Lake is at the north border of image. This image shows fossil fuel CO2 sources according to 2005 EDGAR inventory. Edmonton is the spot close to the middle of south edge with some other major source (as of this time I haven't been able to identify which sort of industry is it caused by) a little to the west. Oil Sands area are the two spots approximately half-way between Edmonton and Lake Athabasca.

Simulation is run for the end of June/beginning of July 2007 (yes, the anthropogenic emissions inventory used is out of date, I will need to address this issue later).

Anthropogenic emissions transport

Click on the image below to open mpg file showing animation of anthropogenic emissions transport. This simulation domain is specific, because it contains just a few of large point sources with much higher anthropogenic fluxes than the rest of area.

Transport of anthropogenic emissions

Transport of anthropogenic emissions from EDGAR. Click for mpg animation.

Biogenic fluxes - VPRM

VPRM input fields have been prepared by VPRM preprocessor for the year 2007. All indices (EVI, LSWI, VEGFRA) appear to be nonzero and hopefully reasonable. This is the CO2 respiration (click to open the animation - size cca 10MB).

Biospheric CO2 release - surface fluxes

Vegetation respiration - VPRM model. Click for mpg animation.

And here is CO2 photosynthesis uptake. All times are in UTC and this is UTC-6 timezone, so first part of animation is empty because there is no photosynthetic activity occuring during the night. The sign of CO2 fluxes is positive here as an easy fix to colour palette problem - photosynthesis is removing CO2 from the atmosphere.
Biospheric CO2 intake - surface fluxes

Gross ecosystem exchange - VPRM model. Click for mpg animation.

Total CO2 concentrations

Here is everything together - i.e. the effects of respiration and photosynthesis fluxes, anthropogenic emissions, boundary conditions and atmospheric transport. Following animation shows expected CO2 concentrations according to WRF-VPRM model. Displayed here is the average of the lowest 9 layers of model to mitigate the effects of vertical transport.

Tropospheric CO2 concentrations.

Tropospheric CO2 concentrations - Click for mpg animation.


This work is supported by Proof of Principle Fund provided by the Canada School of Energy and Environment ...

Bulk of the work presented here uses WRF-VPRM model (an extension of WRF-Chem model) developed at Max-Planck Institute for Biogeochemistry, Germany. Continuous feedback and support from dr. Christoph Gerbig and his team is gratefully acknowledged.


Last change: 13 April 2010
by Pavel Jurus.