In Nepal’s Middle Hills, community-managed forests have successfully reversed deforestation, but they are now unintentionally contributing to water insecurity. Afforestation has heavily favored Pinus roxburghii, a fast-growing conifer with high year-round evapotranspiration and low infiltration capacity, significantly reducing groundwater recharge. As pine offers limited economic value, forest users increasingly shift to Sal (Shorea robusta) forests, valued for timber and compostable leaf litter. This shift concentrates human activity—such as litter collection, grazing, and trampling—around Sal patches, causing surface compaction and further reducing infiltration. Combined with unplanned road construction that disrupts natural flow paths, these disturbances have degraded upland recharge zones. Once crucial for replenishing groundwater, these uplands are now losing their recharge capacity, leading to measurable declines in groundwater storage and drying of springs in foothill and riparian zones that once flowed year-round.
The consequences are widespread and socio-ecologically severe. Rural and Indigenous communities relying on spring-fed systems for drinking water, irrigation, and livestock now face escalating dry-season scarcity. Women and elderly members of marginalized groups bear the greatest burden, while increasing outmigration to urban centers exacerbates inequality. Yet forest governance remains focused on canopy cover and carbon sequestration, often overlooking essential hydrological processes like infiltration, baseflow, and subsurface storage.
The continued decline in groundwater recharge also raises long-term concerns about shallow aquifer sustainability and overall water security. This situation is further complicated by a lack of reliable, long-term ground-based hydrometeorological data—many precipitation, temperature, and stream discharge records are missing or incomplete due to sensor failure—making it difficult to calibrate ecohydrological models and to design informed forest and water policies.
Fortunately, space-based technologies provide a powerful solution. Remote sensing allows for long-term monitoring of vegetation, precipitation, soil moisture, and terrain, revealing the drivers of spring decline. When paired with ecohydrological modeling and community knowledge, these tools can guide forest management strategies that restore groundwater recharge and help achieve SDG targets 6, 13, and 15.
Goals and milestones
The main goal of this research is to assess and mitigate ecohydrological trade-offs in Nepal’s Middle Hills caused by unscientific forest expansion under community forestry (CF). While CF has successfully increased forest cover, it has often overlooked hydrological impacts—particularly where high water-use species like pine have been planted without considering water balance consequences. This has led to declining baseflows, reduced groundwater recharge, and increased dry-season water stress.
A key focus is to bridge the gap between Indigenous forest management practices and scientific understanding of forest-water interactions. By integrating Regional Hydro-Ecological Simulation System (RHESSys), ecohydrological modeling, satellite remote sensing, and community-level knowledge, the project aims to reveal how forest type, topography, and land use influence spring recharge zones, groundwater dynamics, and soil moisture retention.
Research has shown that nearly 70 per cent of the springs in the region are degrading, threatening long-term water security. One of the critical goals of this research is to identify vulnerable and resilient spring zones—and ultimately support the rebirth of these springs through improved forest and land-use strategies.
Short-term milestones include generating high-resolution maps of vegetation phenology, evapotranspiration, and groundwater storage (1985–2025), and validating RHESSys outputs with both field data and satellite products. In the mid-term, the study will identify groundwater-rich zones for future water-resilient settlements, simulate climate scenarios, and collaborate with local stakeholders. The long-term objective is to promote scientifically informed, community-adapted forest governance that enhances both ecological and water resilience across Nepal’s Middle Hills.
Has this problem been acknowledged in the past?
Yes, this problem has been acknowledged by the government, NGOs, and international partners.
Can this challenge be solved using space technologies and data?
This challenge can be effectively addressed using space-based technologies such as Global Precipitation Measurement (GPM), Moderate Resolution Imaging Spectroradiometer (MODIS), Sentinel-2, Landsat, Gravity Recovery and Climate Experiment (GRACE), and Shuttle Radar Topography Mission (SRTM), which provide critical data on precipitation, vegetation, groundwater, and terrain—enabling RHESSys to guide water-resilient forest management in Nepal’s data-scarce, complex mid-hill watersheds.
Parameters and variables to be mapped by a solution
To address the goals and simulate ecohydrological processes in Nepal’s Middle Hills, several key variables are being mapped. I have successfully completed about 50 per cent of this work, while expert support is needed to refine or complete the remaining variables—particularly in addressing technical limitations related to satellite data and downscaling.
- A Land Use / Land Cover (LULC) analysis from 1985 to the present uses Landsat imagery and the Global Land Cover dataset developed by the Chinese Academy of Sciences. This dataset, which includes 35 classified land cover types, enables the differentiation of coniferous and broadleaf forests—critical for simulating forest-water interactions and green-blue water partitioning.
- The Normalised Difference Vegetation Index (NDVI), the Enhanced Vegetation Index (EVI), Leaf Area Index (LAI), and Evapotranspiration (ET) have been mapped. These products are affected by persistent cloud cover and atmospheric noise and expert guidance to apply cloud-removal and correction techniques to improve data accuracy is needed.
- Calculating the Topographic Wetness Index (TWI), slope, aspect, and other terrain features from Digital Elevation Models (DEMs) to delineate recharge-prone areas and spring emergence zones was started. Initial steps toward identifying spring vulnerability and resilience hotspots, as well as settlement suitability zones—based on groundwater access and socio-hydrological factors—are underway and need to be refined with expert support.
- Spring vulnerability and resilience hotspots, integrating remote sensing, modeled water balance, and field survey data.
- Settlement suitability zones, determined based on groundwater availability, hydrological resilience, and socio-ecological stability.
The following variables have yet to be accurately mapped and require technical assistance:
- Mapping vegetation phenology (leaf-on and leaf-off periods),
- Downscaling NDVI, EVI, LAI, and ET to a 30 m resolution using terrain and landscape metrics - is a key objective.
- Accurate mapping of soil moisture from platforms such as Soil Moisture Active Passive (SMAP) is necessary to support ecohydrological model inputs and validation.
Assessing saturation deficit, water table depth, and groundwater storage using RHESSys simulations, supplemented by Gravity Recovery and Climate Experiment (GRACE) satellite data is needed. Support in downscaling GRACE data to the watershed level for meaningful hydrological analysis is a necessary step.
One of the most significant limitations is the lack of high-quality, long-term meteorological data, including precipitation, temperature (maximum and minimum), and stream discharge, due to sparse networks and sensor failures. This gap significantly complicates model calibration and policy-relevant research. Therefore, learning to map and apply satellite-based precipitation and climate variables is essential to enhance model robustness and simulate ecohydrological processes more accurately.
Trends identified using remote sensing
Using National Oceanic and Atmospheric Administration (NOAA) Advanced Very High Resolution Radiometer (AVHRR) (1985–2002) and MODIS (2001–2024):
- NDVI, EVI, and LAI increased by 12–18 per cent, indicating strong greening.
- Greening is mostly from pine species, which are high water users.
- MODIS ET data shows a 38 per cent rise in evapotranspiration, reducing water available for springs and communities.
Insights based on modelling
RHESSys simulations show that:
- Broadleaf forests, primarily Sal trees, reduce peak monsoon flows and improve dry-season baseflows due to lower ET.
- Pine forests increase year-round transpiration, lowering groundwater recharge.
- Mixed-species forests offer balanced hydrology and better ecological resilience.
These results highlight how species composition directly impacts water availability, especially in dry seasons.
Hydrological and groundwater trends identified through modeling
From 1990 to 2022:
- Streamflow declined by ~22 mm/year, especially in the dry season.
- Groundwater storage dropped due to high canopy interception and ET.
- RHESSys outputs matched MODIS LAI, validating model accuracy.
MODIS data from 2003 to 2022 was used to analyse trends in comparison with RHESSys-generated LAI. For the period before 2003, NOAA data was used, although it has a course spatial resolution of 5,560 m.
Spring and recharge mapping from modeling need to be integrated with remote sensing as well as indigenous people perception.
- Lower-elevation broadleaf zones, primarily dominated by Sal trees, have shallow water tables.
- Pine-dominated uplands show deeper groundwater and water stress.
- Mapped spring hotspots and recharge zones support future settlement planning and water security strategies.
Expected timeframe to develop a solution
6 months (modeling and strategy development), 4–5 years (results on the ground such as spring revival, baseflow improvement, and recharge enhancement).
Potential consequences if no action happens
- Failure to act will deepen the water crisis across Nepal’s Middle Hills. Springs—already drying in 60% of upland areas—will disappear further, triggering migration, agricultural abandonment, and social instability. Marginalized communities, particularly women and lower-caste groups, will face increased hardship.
- Forest degradation and monoculture expansion will reduce infiltration and elevate flash flood risk. Urban centers like Kathmandu, dependent on mid-hill water sources, will face supply disruptions. Ecosystem services will erode, compromising biodiversity and climate resilience.
- The cascading impacts will threaten multiple SDGs—not just water (6), climate (13), and terrestrial ecosystems (15), but also food security (2), health (3), poverty reduction (1), and gender equity (5).
What are additional physical requirements for a solution?
Securing water resources in Nepal’s mid-hill watersheds requires an integrated approach combining space-based data, field engagement, and ecohydrological modelling. Key inputs include high-resolution satellite datasets (GPM, SM2RAIN-ASCAT, APHRODITE, MODIS, GRACE, Sentinel-2, and DEMs) to monitor precipitation, vegetation dynamics, evapotranspiration, and groundwater storage. These feed into the RHESSys model to simulate long-term changes in the water budget. Field-based efforts—such as household and Community Forest User Groups (CFUG) surveys, participatory mapping, and basic hydrological monitoring—are essential to ensure scientific insights reflect local realities. Successful implementation also depends on physical access to watersheds and strong collaboration with local institutions, professionals, stakeholders, and NGOs to co-develop forest-water strategies that restore spring flow, enhance groundwater recharge, and build long-term water security.
These ecohydrological stresses are intensified by unsustainable forest practices such as litter removal, trampling, and understory clearance—especially in critical recharge areas. As spring-fed systems degrade, communities face agricultural decline, rising food insecurity, and forced migration. The burden falls especially on women and marginalized castes who must travel further for water.
Despite increasing forest cover, CF governance frequently overlooks hydrological consequences, focusing on forest density or carbon stocks instead. These regions also lack adequate hydrological monitoring, with sparse data on rainfall, discharge, and groundwater recharge. Addressing this issue requires an integrated framework that connects forest composition, land use, and climate impacts, while also accounting for socio-institutional dynamics and equity.
What exactly do I aim to do with the support of the Space4Water community?
The first step I aim to take with the support of the Space4Water community is to build a robust understanding of the most suitable remote sensing datasets and tools that can effectively address the challenges posed by cloud-prone, topographically complex, and data-scarce regions like Nepal’s Middle Hills. Although MODIS datasets have been instrumental, they are often affected by persistent cloud cover and coarse spatial resolution, making it difficult to obtain reliable vegetation and hydrology-related insights for smaller, fragmented watersheds.
A particular focus of my research is on vegetation phenology (leaf-on and leaf-off timing), which plays a critical role in determining seasonal evapotranspiration and soil moisture dynamics. The MODIS phenological and vegetation products, though informative, operate at 250m to 500m resolution, which does not suffice for the highly heterogeneous vegetation mosaic in Nepal’s mid-hills. Therefore, I plan to downscale these datasets to 30m resolution by coupling them with Landsat data, digital elevation models (DEMs), slope, aspect, and other topographic variables. This approach will enable the differentiation of key vegetation types—particularly pine vs. broadleaf forests, which differ significantly in water-use efficiency.
Since conifer plantations that had been carried out without scientific analysis have been linked to the drying springs and reduced groundwater recharge, it is crucial for sustainable forest and water resource planning in community-managed landscapes to understand species-level water use dynamics .
Additionally, I aim to engage with stakeholders, experts, and young professionals working in similar mountainous or Indigenous regions, to co-develop location-appropriate algorithms and share insights on forest hydrology, remote sensing integration, and community water resilience. The Space4Water platform offers a unique opportunity for such interdisciplinary exchange, which is essential for designing effective, science-based interventions grounded in local realities.
In which areas do I need support?
To realize the above objectives, I need targeted support in the following key areas but any small help and suggestions to these will be great.
1. Access to high-resolution and cloud-free satellite data: Given Nepal’s frequent cloud cover and steep terrain, I need guidance on accessing and processing datasets like Sentinel-1 SAR (for cloud-penetrating vegetation and soil moisture mapping), Sentinel-2, and Landsat, as well as cloud-free composites and advanced cloud-removal algorithms.
2. Downscaling of remote sensing products: Support is required to downscale MODIS-based variables such as NDVI, EVI, LAI, and phenological timing using auxiliary data like DEM, slope, aspect, and land cover classification, thereby enabling more granular forest-type discrimination.
3. Integration of hydrological and remote sensing datasets: I plan to combine MODIS Evapotranspiration, SMAP soil moisture, and GRACE groundwater storage anomalies with outputs from the RHESSys ecohydrological model. Guidance on effectively integrating these datasets to cross-validate model results and monitor vegetation–water interactions is vital.
4. Mapping groundwater recharge and spring hotspots: I seek technical support for mapping spring-sustaining terrain features (e.g., toe slopes, concave landforms, riparian corridors) using remote sensing and RHESSys-derived saturation metrics, to inform future settlement planning and water security interventions.
5. Knowledge exchange and capacity building: Lastly, connecting with community hydrology researchers and practitioners working in Latin America, Africa, or Southeast Asia could greatly enrich our mutual understanding of groundwater-spring-forest interactions in similar socio-ecological contexts.
Periods of interest
Our primary goal is to assess the impact of community forest expansion on spring and groundwater dynamics, focusing on both pre- and post-CF eras. The study spans from 1985 to 2025, segmented as follows:
1985–1992 (Baseline period): Historical land cover and forest structure derived from Landsat and MODIS data.
1993–2022 (Community forestry implementation): Analysis of vegetation growth, forest composition changes, spring discharge trends, and streamflow responses under CF.
2023–2025 (Current assessment): Field surveys, interviews with Indigenous and local communities, and continued RHESSys model refinement.
This temporal window captures how unregulated plantation and species shifts—often lacking hydrological consideration—have affected spring discharge, soil moisture, and subsurface storage. Many communities report vanishing springs and declining streamflow, prompting seasonal migration in search of water.
While the core analysis runs through 2025, these results will be used to project future scenarios under Coupled Model Intercomparison Project Phase 6 (CMIP6) SSPs (2025–2050), enabling us to simulate potential consequences of climate change and forest management strategies—and thus inform timely, preventive policy responses.
