Editing IPCC:AR6/SRCCL/Chapter-4 (section)

==== 4.4.1.1 Changes in water erosion risk due to precipitation changes ====

<div id="section-4-4-1-1-changes-in-water-erosion-risk-due-to-precipitation-changes-block-1"></div>

The hydrological cycle is intensifying with increasing warming of the atmosphere. The intensification means that the number of heavy rainfall events is increasing, while the total number of rainfall events tends to decrease (Trenberth 2011 <sup>[[#fn:r541|541]]</sup> ; Li and Fang 2016 <sup>[[#fn:r542|542]]</sup> ; Kendon et al. 2014 <sup>[[#fn:r543|543]]</sup> ; Guerreiro et al. 2018 <sup>[[#fn:r544|544]]</sup> ; Burt et al. 2016a <sup>[[#fn:r545|545]]</sup> ; Westra et al. 2014 <sup>[[#fn:r546|546]]</sup> ; Pendergrass and Knutti 2018 <sup>[[#fn:r547|547]]</sup> ) ( ''robust evidence, high agreement'' ). Modelling of the changes in land degradation that are a result of climate change alone is hard because of the importance of local contextual factors. As shown above, actual erosion rate is extremely dependent on local conditions, primarily vegetation cover and topography (García-Ruiz et al. 2015 <sup>[[#fn:r548|548]]</sup> ). Nevertheless, modelling of soil erosion risks has advanced substantially in recent decades, and such studies are indicative of future changes in the risk of soil erosion, while actual erosion rates will still primarily be determined by land management. In a review article, Li and Fang (2016) <sup>[[#fn:r549|549]]</sup> summarised 205 representative modelling studies around the world where erosion models were used in combination with downscaled climate models to assess future (between 2030 to 2100) erosion rates. The meta-study by Li and Fang, where possible, considered climate change in terms of temperature increase and changing rainfall regimes and their impacts on vegetation and soils. Almost all of the sites had current soil loss rates above 1 t ha–1 (assumed to be the upper limit for acceptable soil erosion in Europe) and 136 out of 205 studies predicted increased soil erosion rates. The percentage increase in erosion rates varied between 1.2% to as much as over 1600%, whereas 49 out of 205 studies projected more than 50% increase. Projected soil erosion rates varied substantially between studies because the important of local factors, hence climate change impacts on soil erosion, should preferably be assessed at the local to regional scale, rather than the global (Li and Fang 2016 <sup>[[#fn:r550|550]]</sup> ).

Mesoscale convective systems (MCS), typically thunder storms, have increased markedly in the last three to four decades in the USA and Australia and they are projected to increase substantially (Prein et al. 2017 <sup>[[#fn:r551|551]]</sup> ). Using a climate model with the ability to represent MCS, Prein and colleagues were able to predict future increases in frequency, intensity and size of such weather systems. Findings include the 30% decrease in number of MCS of &lt;40 mm h <sup>-1</sup> , but a sharp increase of 380% in the number of extreme precipitation events of &gt;90 mm h <sup>–1</sup> over the North American continent. The combined effect of increasing precipitation intensity and increasing size of the weather systems implies that the total amount of precipitation from these weather systems is expected to increase by up to 80% (Prein et al. 2017 <sup>[[#fn:r552|552]]</sup> ), which will substantially increase the risk of land degradation in terms of landslides, extreme erosion events, flashfloods, and so on.

The potential impacts of climate change on soil erosion can be assessed by modelling the projected changes in particular variables of climate change known to cause erosion, such as erosivity of rainfall. A study of the conterminous United States based on three climate models and three scenarios (A2, A1B, and B1) found that rainfall erosivity will increase in all scenarios, even if there are large spatial differences – a strong increase in the north-east and north-west, and either weak or inconsistent trends in the south-west and mid-west (Segura et al. 2014 <sup>[[#fn:r553|553]]</sup> ).

In a study of how climate change will impact on future soil erosion processes in the Himalayas, Gupta and Kumar (2017) <sup>[[#fn:r554|554]]</sup> estimated that soil erosion will increase by about 27% in the near term (2020s) and 22% in the medium term (2080s), with little difference between scenarios. A study from Northern Thailand estimated that erosivity will increase by 5% in the near term (2020s) and 14% in the medium term (2080s), which would result in a similar increase of soil erosion, all other factors being constant (Plangoen and Babel 2014 <sup>[[#fn:r555|555]]</sup> ). Observed rainfall erosivity has increased significantly in the lower Niger Basin (Nigeria) and is predicted to increase further based on statistical downscaling of four General Circulation Models (GCM) scenarios, with an estimated increase of 14%, 19% and 24% for the 2030s, 2050s, and 2070s respectively (Amanambu et al. 2019 <sup>[[#fn:r556|556]]</sup> ).

Many studies from around the world where statistical downscaling of GCM results have been used in combination with process-based erosion models show a consistent trend of increasing soil erosion.

Using a comparative approach, Serpa et al. (2015) <sup>[[#fn:r557|557]]</sup> studied two Mediterranean catchments (one dry and one humid) using a spatially explicit hydrological model – soil and water assessment tool (SWAT) – in combination with land-use and climate scenarios for 2071–2100. Climate change projections showed, on the one hand, decreased rainfall and streamflow for both catchments, whereas sediment export decreased only for the humid catchment; projected land-use change, from traditional to more profitable, on the other hand, resulted in increase in streamflow. The combined effect of climate and land-use change resulted in reduced sediment export for the humid catchment (–29% for A1B; –22% for B1) and increased sediment export for the dry catchment (+222% for A1B; +5% for B1). Similar methods have been used elsewhere, also showing the dominant effect of land-use/land cover for runoff and soil erosion (Neupane and Kumar 2015 <sup>[[#fn:r558|558]]</sup> ).

A study of future erosion rates in Northern Ireland, using a spatially explicit erosion model in combination with downscaled climate projections (with and without sub-daily rainfall intensity changes), showed that erosion rates without land management changes would decrease by the 2020s, 2050s and 2100s, irrespective of changes in intensity, mainly as a result of a general decline in rainfall (Mullan et al. 2012 <sup>[[#fn:r559|559]]</sup> ). When land management scenarios were added to the modelling, the erosion rates started to vary dramatically for all three time periods, ranging from a decrease of 100% for no-till land use, to an increase of 3621% for row crops under annual tillage and sub-days intensity changes (Mullan et al. 2012 <sup>[[#fn:r560|560]]</sup> ). Again, it shows how crucial land management is for addressing soil erosion, and the important role of rainfall intensity changes.

There is a large body of literature based on modelling future land degradation due to soil erosion concluding that, in spite of the increasing trend of erosive power of rainfall, ( ''medium evidence, high agreement'' ) land degradation is primarily determined by land management ( ''very high confidence'' ).

<div id="section-4-4-1-2-climate-induced-vegetation-changes-implications-for-land-degradation"></div>

<span id="climate-induced-vegetation-changes-implications-for-land-degradation"></span>