Pan-tropical modelling of the effects of tree cover growth on water quantity and regulation services

Student thesis: Doctoral ThesisDoctor of Philosophy

Abstract

Forest restoration and reforestation projects are multiplying in the tropics and globally, but most of them are designed for climate change mitigation and often neglect to critically examine the hydrological services provided by forests. Fur-thermore, the impact of large-scale afforestation on water quantity and regulation in the tropics is not fully known. Lacking specifically, is the knowledge of where and when forest increase can lead to more positive or more negative hydrological outcomes, which is at the centre of the "infiltration-evapotranspiration trade-off” debate. Equally, there are no hydrological models suitable to address this gap on large spatial scales and with publicly available datasets that have good coverage in the tropics.

The aim of this study was to address these gaps in several ways. A new spatially distributed hydrological model, DTRIE, was developed to adequately represent the effects of tree cover on surface and subsurface flows on large spatial scales. DTRIE operates on a monthly time step to model long-term changes to annual and seasonal water availability. Second, the study produced a novel, realistic scenario for tree cover growth in the tropics which is constrained by environmen-tal and socio-economic data. Third, a data-driven incremental growth model to estimate fractional tree cover at multiple management-relevant time steps was developed.

The key findings from this study are as follows: DTRIE performed acceptably when compared with estimates from other published models and datasets, such as evapotranspiration data from MODIS or GLEAM, or monthly stream discharge from the Global Runoff Data Centre (GRDC).

Across the tropics, 812 million hectares of area of opportunity for tree cover growth were identified, most of which is located in Central and South America, followed by Africa, and Asia and Oceania. The tree cover growth model projected an increase in tropical tree cover from 25.8 per cent (current/ in 2018) to 31.3 per cent (in 2118). Because of low tree cover in 2018, tropical Africa saw the highest relative increase.

When applied to the scenario tree cover conditions, DTRIE results supported the consensus that annual water quantity declines with more tree cover, except where cloud forests are present. A new finding was that the sensitivity of water balance to tree cover change, i.e., the change in water balance per change in unit tree cover, averaged over larger tropical basins, was generally low. The max-imum change in water balance per unit change in tree cover was 0.6 mm, but 90 per cent of the values were below 0.25 mm. The sensitivities of AET (actual evapotranspiration) and infiltration to forest increase were compared. Areas where increased infiltration exceeded increased AET after tree growth were found to have a higher baseline tree cover and a higher aridity index (meaning they were more humid). These areas have the potential to see increased dry sea-son flows because less water leaves the system as AET.

The intra-annual variability of streamflows in several tropical dam watersheds in-creased more steeply with tree cover growth when watersheds were more humid and more area was under restoration. In some watersheds dry season flows in-creased despite an overall more uneven distribution throughout the year.

These findings are relevant in the seasonally dry tropics, and especially where streamflow regulation is important. For example, in regions with high dependency on irrigated agriculture but low-capacity dam infrastructure upstream forestation may be beneficial. Here, increases in dry season flows may outweigh the ex-pected decreases in annual water quantity through forestation. Conversely, since the reduction of water quantity is almost a given, areas where regulation is not an issue but overall quantity is important may not want to sacrifice water quantity for an increase in tree cover.
Date of Award1 May 2024
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorJane Catford (Supervisor) & Mark Mulligan (Supervisor)

Cite this

'