The role of the hydrological cycle in forest ecosystems : flow path, nutrient cycling and water-carbon interaction

Abstract: Forest ecosystems, covering over a third of land on the Earth, play a significant role in the global hydrological cycle, and influence soil erosion and climate change. However, the distribution, movements, quality of water, and hydrological processes in forested ecosystems are not well understood yet. This thesis aims to improve our understanding of the interaction between forest ecosystems and water cycle from the perspective of flow path, nutrient cycling, and carbon – water interactions. Flow path is particularly important for the study of water storage and distribution, and solute transport and attenuation. However, in dense forest areas, flow path is usually hard to detect from terrain models due to large noise in elevation data, e.g. large sinks. Spurious sinks hinder water flowing downslope and thus likely result in unrealistic flow path estimation. An algorithm that can tackle spurious sinks without altering elevation was proposed and shown to be able to estimate flow path more accurately than traditional methods for different terrain forms. Besides the problem of flow path estimation, the evaluation of flow path estimation has usually been done for the whole catchment, ignoring the variability of the disagreement between estimated flow path and observations among different land cover, soil type, and slopes within a catchment. A number of culverts investigated in fields have thus been used and taken as observations of stream locations for the assessment of flow path evaluation. The results showed that the uncertainty of flow path estimation is strongly related to soil hydraulic productivity, vegetation cover, and slope. Furthermore, nutrient cycling can significantly affect the quality of water. Water is observed getting browner in (sub)arctic regions due to elevated concentration of dissolved organic carbon (DOC). However, it is still not well known how catchment morphometric, e.g. hydrologic connectivity, could affect the distribution and transportation of DOC from terrestrial systems to streams. A systematic analysis of the relationship between catchment morphometric and DOC concentration was thus carried out, and the results showed that smaller catchment size, shorter flow length, and younger catchments tend to have higher DOC concentrations. Moreover, remote sensing observations have revealed that Amazon rainforests are resilient to droughts and could maintain photosynthesis productivity during dry season. Light, leaf phenology etc. have been reported to drive the seasonality of vegetation productivity in tropical forests. Plant use of groundwater with deep roots could sustain evapotranspiration during the dry season and thus could drive the dry season greening. However, the role of groundwater in the resilience of tropical forests to drought is not well studied yet. A global dynamic vegetation model, i.e. LPJ-GUESS model, was implemented to simulate water and carbon fluxes for tropical forests. Model simulations with groundwater introduced were found to be able to reproduce the seasonality of photosynthesis productivity of tropical forests. In addition, model simulations showed that groundwater substantially sustains tropical forest productivity, especially for wet and seasonally dry areas if future climate is getting drier. However, it seems that groundwater doesn’t make large differences for forest productivity when future climate is getting wetter.