Freshwater biodiversity is in a state of crisis with an annual decline rate of 3.9% compared to 1.1% decline in terrestrial biodiversity (Living Planet Index). The actual decline rate may be higher than estimated as many of the species are data deficient or not yet assessed or became extinct before assessment. We have far surpassed the sustainable or natural limit of reductions in populations of freshwater species. With an unprecedented increase in global energy and water demand, planetaries boundaries for sustainable use of freshwater may be overstepped in the near future, as presently visible at a local scale. A staggering decline in freshwater biodiversity is a first warning signal of a looming global water crisis. With continued pressures like river fragmentation, flow regulation, overextraction of surface or groundwater, pollution, invasive species and climate change, on freshwater habitats, this crisis may shift one level higher and impact human species severely. Environmental flows forms the link between ecological health of a river and the ecosystem services we derive from it. It can be a great tool to achieve twin objective of maintaining the ecosystem integrity of freshwater habitats and deciding trade-offs for ensuring sustainable water management in a river basin. This study focuses on developing a holistic methodology for preliminary assessment of ecosystem integrity or ecological health of a river basin, which can be easily adapted to other river basins using open source global datasets. The proposed methodology was applied to Kaveri basin to test its applicability and identify the limitations of available global datasets. A widely accepted regional environmental flow assessment framework, ELOHA (Ecological Limits of Hydrological Alteration) was adapted by using global datasets. A global river classification dataset was used to identify the river classes in Kaveri basin. Monthly hydrological alteration in magnitude at the location of gauge stations was calculated, using PCR-GLOBWB data as reference for natural flow conditions, in absence of records for natural flow in a highly modified Kaveri basin. An ecosystem integrity indicator framework was developed to assess the hydrologic, geomorphic and ecological modifications in the river basin. Indicators grouped under four main categories - Connectivity status, Land Use, Biodiversity, Water Quality were adopted using exiting global datasets and values for all the sub-basins of Kaveri basin were estimated. Finally an attempt to derive flow alteration-ecological response was made. Threatened fish species percentage, quantified using IUCN spatial dataset, showed an increase in value with increase in alteration in flow magnitude. No clear relationship was observed when data for other taxonomic groups like plants, molluscs, odonata, shrimps and crabs were used. Hence, species of concern (IUCN red list category - CR, EN,VU) data can be useful in deriving preliminary flow alteration-ecological response relationship. An attempt to find linkage between flow alteration and floodplain gross primary productivity was also made. In dry season an inverse relationship was observed at few gauge stations but in general other climatic factors like rainfall and evapotranspiration had greater influence on gross primary productivity. Impact on gross primary productivity due to flow alteration could not be isolated using existing datasets because of coarse resolution.

The Morgan-Morgan-Finney (MMF) model is identified as the most appropriate hillslope soil erosion model for high resolution global application with the objective of identifying relative erosion risk areas and modelling the effect of land cover changes on erosion at a catchment scale. This is justified by its simple model structure, low input requirements, semi-empirical basis and distributed application. However, transferability to global application is hindered by its reliance on empirical data for input requirements. This paper proposes methodologies of generating input data based on remotely sensed products which improves the spatial and temporal accuracy of the input rasters. The MMF model is further redesigned by coupling it with the IHE Delft Institute for Water Education inhouse hydrological model, WaterPix, to replace runoff calculations in order to improve its treatment of infiltration and by applying the model in monthly timesteps to analyse erosional differences within the seasonal crop calendar. The redesigned MMF model algorithm is coded in python and applied over a headwater in the Ganga basin located in the Madhya Pradesh state of India. The model produces realistic erosion rates and distributions and provides additional information on the spatial and temporal variation of erosion in the study area. However, validation of the redesigned model with field data needs to be prioritized before it is utilized.