This is a page about the

         Fluxes from Flooded Vegetation
                      Discussion Group

coordinated by Toby Marthews at UKCEH (email me if you are interested in joining the email list for this). Please see also our blog.

Schedule of meetings:
     Meeting 1: 9-NOV-21. Presentation by Ross Morrison about UKCEH flux tower data.
     Meeting 2: 8-DEC-21. Presentation by Toby Marthews about flooded vegetation and another by Nic Gedney about her upcoming paper.
     Meeting 3: 8-FEB-22. Presentation by Noah Smith and Toby Marthews on soil supersaturation.
     Meeting 4: 8-MAR-22 Presentation by Anthony Schrapffer about ORCHIDEE's new Floodplains Scheme and results from the Pantanal.
     Meeting 5: 12-APR-2022. Presentation by Athanasios Paschalis about his group's current work on the Tethys-Chloris wetland model.
     Meeting 6: 7-JUN-2022. Presentation on river routing/inundation modelling by Toby.

   ** Discussion Group closed - there will be no further meetings! **

Here are the slides I presented in September 2021 on this topic and here are the slides I presented on 12th October 2021 at our (Zoom) kick-off meeting. I am trying to work out how to address the issues on this topic listed below.

• What new observational data do we need?
• What new theories do we need?
• What approaches do various models currently take to quantifying fluxes from flooded vegetation?

Toby Marthews, October 2021.

PS. This group was never a JULES Process Evaluation Group (JPEG, e.g. search for "JPEG" here) because it was not a JULES-specific discussion.

Global wetlands and other inundated areas: why are they important?

• Wetlands are the largest single natural source of atmospheric methane (CH4) and form a key link between the water and carbon cycles at regional scales.
• Wetland ecosystems mediate large land-atmosphere transfers of heat, water and trace gases at both local and regional scales.
• Global warming resulting from a doubling of atmospheric CO2 concentrations will likely lead to a 78% increase in wetland emissions of CH4, most of which is expected to come from tropical regions (Shindell et al. 2004, Gedney et al. 2019).

Wetlands are generally considered peripheral land use units because for the majority of humans they are: you can't build on them and they are too wet for agriculture, so they are ignored. It is only relatively recently that we are realising they play a key role in many ecosystems and the global climate system (e.g. see WWT). However, for me it is climate change that really makes wetlands crucial: that 78% figure above is enormous and we must find a way to mitigate this.

The lack of robust information on the ways in which wetlands fluctuate and function (not just tropical ones) is hampering progress in predicting the effects of global environmental change. We urgently need a better understanding of how wetlands function in the Earth system.

Flooded vegetation is not generally treated explicitly in land surface models. In the JULES code, for example, ice and lake areas on land are assumed to have zero vegetation (and rivers have no width). Routines exist within JULES for specifying irrigated area and predicting inundated area, but these areas are not dynamically linked to the vegetation tiles within the gridcell, meaning that 'flooded vegetation' does not currently exist as a concept within JULES.

People have been thinking about this for a while, e.g. Dadson et al. (2010) "Vegetated zones can and do become flooded, in proportion to their original surface cover fractions. At present we have insufficient information on the detailed biophysical response of vegetation to flooding to model this process explicitly, but we note that potential improvement for future work.". Also, Miguez-Macho & Fan (2012) had a section Evaporation From Floodwater Surface (although no mention of the flooded vegetation).

Think about these points:

• Wetlands, which are areas of flooded vegetation, cover an enormous area of land on all continents: e.g. see map of US (left) from here (and if England had not drained so much of its land, I believe the proportion there would be pretty similar).
  
• Swamp forests are globally extensive (e.g. the Congo Swamp Forests cover an area of 150,000 km2, or equivalent to England + Wales), but JULES effectively assumes their dynamics are identical to the terra firme forest types that surround them.
• Fluxes from flooded vegetation are known to be different, (e.g. CO2, ET) and recent work has shown how soil moisture status affects vegetation in dry soils (Harper et al. 2021).
• Near large rivers, vegetation can change enormously (most famously, Brazil's igapó and várzea forests, but similar variation can be seen in other continents). However, species overlap is often very high so I would be against defining a new PFT for swamp forest (at least for now): we need first to find some way of inserting the physical processes by which trees 'know' they are inundated and therefore can behave differently.
• Flooded vegetation is an intermediate land cover state between terra firme forest and lake / open water areas. JULES does include some routines that handle fluxes from lakes (FLake: see Rooney & Jones 2009 and this RMED ticket and Rooney & Bornemann 2013 and Helen Johnson's talk 2021), therefore perhaps a first guess would be to assume that fluxes from flooded vegetation type A are always simply halfway between the flux from terra firme A and the flux from a lake at that location? I have no data that would support this (and I'm open to the suggestion e.g. that the relationship is hump-backed with flooded vegetation having higher fluxes than terra firme or lake), but implementing something like this would probably be a good exercise to find out what is possible.