Revision of methane and carbon dioxide emissions from inland waters in India

Substantial greenhouse gas (GHG) emissions including methane (CH₄) and carbon dioxide (CO₂) from inland waters have captured international concern (Barros et al., 2011; Bastviken et al., 2011; Raymond et al., 2013; Panneer Selvam et al., 2014). Global freshwaters are consistently oversaturated with CO₂ in contrast to atmospheric level resulting in a significant source of atmospheric CO₂ (Raymond et al., 2013). Green house gas emissions from global inland waters are reported to be 0.65 Pg of C (CO₂ eq) yr⁻¹ as CH₄ (Bastviken et al., 2011) and 1.2–2.1 Pg C yr⁻¹ as CO₂ (Raymond et al., 2013). However, global estimates are constrained by data paucity and poor coverage of Asia in particular. Recently Panneer Selvam et al. (2014) showed an impressive estimate of GHG emissions from inland waters in India, yielding a substantial revision of GHG emission from global freshwaters. However, the major rivers such as Ganges–Brahmaputra, Indus and Peninsular rivers are not sampled. Hence, rivers were potentially not representative for the whole of India. Moreover, current estimates suffered from data limitation on reservoirs particularly GHG emission from drawdown zone and reservoir downstream (spillways and turbines), where are recognized to be significant carbon emitters (Lima et al., 2008). This should be another major concern, for example, Lima et al. (2008) estimated annual CH₄ emission of 14.2 Tg CH₄ yr⁻¹ from India's large dams (4005 dams are modeled), and this data could be 33.4 Tg CH₄ yr⁻¹ if spillways/turbines are fully considered (Table 1). Of which, 1.1 Tg CH₄ yr⁻¹ was from the reservoir surface, three times higher than the current emission from reservoir/barrage (0.33 Tg CH₄ yr⁻¹; Panneer Selvam et al., 2014).

Here, we identify a gross underestimation of GHG evasion from India with focuses on riverine CO₂ emission and GHG from spillways/turbines, providing a revised and arguably more complete account for the emission of GHG from inland freshwaters of India.

Aqueous partial pressure of CO₂ was consistently higher than atmospheric level (ca. 390 μatm), ranging from 640 to 2700 μatm (Fig. 1), hence leading to a CO₂ flux ranging from 53 to 494 mmol m⁻² day⁻¹ in the major rivers of India. Our estimated area weighted CO₂ flux was on average 163 mmol m⁻² day⁻¹, which was eightfold higher than the current estimate of 20 mmol m⁻² day⁻¹ (Panneer Selvam et al., 2014). Thus, the extrapolated CO₂ emission could be 137.6 Tg CO₂ yr⁻¹ when the water area of 52 600 km² for rivers/streams is considered, implying that Indian rivers/streams contributed an extra ca. 117.8 Tg CO₂ yr⁻¹ to the global CO₂ budget. Our extrapolated flux was 38.7% of CO₂ emission from rivers and streams of the United States (ca. 355.7 ± 117.3 Tg CO₂ yr⁻¹; Butman & Raymond, 2011).

The estimated 5.94 Tg CH₄ yr⁻¹ and 0.71 Tg CO₂ yr⁻¹ could be released through spillways/turbines, thus increasing CO₂ emission to be 3.08 Tg CO₂ yr⁻¹ and CH₄ emission to be 6.27 Tg CH₄ yr⁻¹ by Indian reservoirs. The estimated CH₄ emission was still much lower (ca. 44%) than prior estimate by Lima et al. (2008) (Table 1).

In combination with GHG emission from other aquatic systems in India (Panneer Selvam et al., 2014), we derive that a total of 140.52 Tg CO₂ yr⁻¹ and 8.07 Tg CH₄ yr⁻¹ was released from India's inland waters. Thus, expressed as CO₂ equivalents (eq), 342.3 Tg CO₂ (eq) is being emitted from India's water bodies. This is equal to around 193.4% of the estimated land sink of 177 Tg CO₂ yr⁻¹ in India. If downstream methane emission of 32.4 Tg CH₄ yr⁻¹ is used (Lima et al., 2008), the total GHG emission by India's waters could be 1004 Tg CO₂ (eq) yr⁻¹, 5.7 times greater than the land sink of India.

Regional level inaccuracy as a result of data limitation and bias in data distribution is the major source behind huge uncertainties for global scale estimates, i.e. 0.5-1.8 Pg C-CO₂ yr⁻¹ by global rivers and streams, and 0.32–0.8 Tg C-CO₂ yr⁻¹ for lakes and reservoirs (Barros et al., 2011; Raymond et al., 2013; Table 1). However, by utilizing the updated estimates (see Table 1), we are able to provide a revised estimate of GHG evasion as follows: 1.8 Pg C-CO₂ yr⁻¹ for rivers and streams (Raymond et al., 2013); 0.8 Pg C-CO₂ yr⁻¹ for lakes and reservoirs (Barros et al., 2011); 103.3 Tg CH₄ yr⁻¹ for rivers, lakes, and reservoirs (Bastviken et al., 2011); 100.8 Tg CH₄ yr⁻¹ from reservoirs' spillways and turbines. This corresponds to 4.0 Pg C-CO₂ (eq)/y. Combining rivers, lakes, and reservoirs with wetland provides a total GHG evasion of 8.9 Pg C-CO₂ (eq) yr⁻¹ from all inland waters (2.1 Pg C-CO₂ yr⁻¹ is from wetlands (Aufdenkampe et al., 2011); 177.4–640 Tg CH₄ yr⁻¹ is from global wetlands if we used the averaged CH₄ emission rate in India's wetlands to upscale to global scale, and a surface area of 42.4-152.2 × 10⁵ km² for wetlands is considered). This implies that methane evasion from inland waters (613 Tg CH₄ yr⁻¹) is higher than anthropogenic CH₄ source (ca. 450 Tg CH₄ yr⁻¹), and CO₂ emission from inland waters is 49.5% of global human's carbon source (9.5 Tg C yr⁻¹).

With these revisions, some uncertainty is addressed and enables a more complete, revised estimate of GHGs emissions of some 141.6 Tg CO₂ yr⁻¹ and 9.1 Tg CH₄ yr⁻¹ is released by India's inland waters. This corresponds to 369 Tg CO₂ (eq) yr⁻¹, which is 4.9-fold higher than the current estimate and 2.1 times greater than the estimated land carbon sink of India (Panneer Selvam et al. (2014). To improve on the current quantification of GHGs on a national scale, new and more efforts to measure flux data associated with dams and data through the monsoon season are urgently needed in India and elsewhere in Asia.