Water table depth governs how hard warming pushes peatland CO2 emissions

The two variables cannot be read separately because of oxygen. Peat accumulates in the first place because waterlogging keeps decomposition slow ^[18]; drainage lets air into the profile, and microorganisms then break the organic material down faster, releasing CO2 ^[14]. The measured pattern is what that chemistry would predict: in the annual budgets, emissions rise considerably more with temperature where the water table sits deep ^[8]. The sensitivity in question is the response of one site's own CO2 balance to temperature, not a global quantity.

How much of this transfers to a project depends on design. The paper, published in Nature Communications ^[1], rests on 276 site-years spread over 114 peatlands, which averages about 2.4 years per site ^[15], so the annual analysis leans on contrasts between sites rather than long records at any one of them. The sites are not matched: natural fens and bogs sit in the same dataset as croplands, grasslands and former peat extraction areas ^[4], and the team used explainable machine learning because vegetation and solar radiation also move the fluxes and the relationships are nonlinear ^[5]. A gradient in water table depth across many sites is still not the same measurement as raising the water table at one site and watching what follows.

That is why the second dataset carries more weight than its size suggests. Its 113 site-years of daily fluxes ^[9] amount to about 41 percent as many site-years as the annual set ^[16], but daily data let temperature vary while land use and plant community hold still, so the temperature response is estimated against a nearly fixed background. Nicolas Behrens of the University of Münster, the paper's first author, says it is the first demonstration across many peatland sites that higher water tables reduce the effect of high temperatures on CO2 emissions ^[10]. Two datasets at different time scales pointing the same way is a firmer result than either alone.

The depth numbers turn out to be messier than they first appear. Emissions fall off sharply once a very low water table is brought to within 60 to 75 centimetres of the surface ^[6], while the range the study identifies for minimising emissions is 20 centimetres or shallower ^[7]. Moving from the first to the second is a further 40 to 55 centimetres of rise ^[17], and on the cropland and grassland sites in this dataset ^[4] that is the increment that competes with the land use itself.

This paper stops short of telling you what a rewetting project is worth. The analysis covers CO2 alone, and the authors note that a complete climate balance needs methane and nitrous oxide, whose emissions can also change as a result of rewetting ^[12]. Nothing in the published account addresses crediting, offsets or national inventories ^[19]. Water table depth is the variable a restoration plan can name and defend from this evidence; this paper does not contain a figure for tonnes of CO2-equivalent avoided at a given target depth.