We use WaterWorld (v2.91, hyperuser) to investigate the current water resources, scenarios for land use change and the impacts of conservation scenarios for Kathmandu Nepal. First we define a 1-hectare tile centered on Kathmandu. Then we prepare data and define a region of interest by watersheds of points and search for the Bagmati river amd mask according to rules where ZOI map Bagmati = 1 . We then mask and copy to workspace and then run a baseline simulation. Tree cover in the basin is limited and the catchment is dominated by herbaceous cover associated with cropping. Some other input data can be visualised using the view by links. Baseline water resources Baseline water balance indicates drier conditions in the valley than the surrounding hillslopes with particularly high inputs on south west facing slopes. This largely reflects the distribution of wind-driven rainfall. The inputs button can be used to check the inputs to this map for particular pixels. Actual evapotranspiration is highest for the forested zones in the north. Inputs from snowmelt in this basin are zero but total fog inputs can be 300mm/yr in the forested mountains, though fog contributions to runoff are highest in the drier lowlands. For the river courses in the lowlands fog contributions are rarely greater than 10% of flow. The human footprint on water quality is high throughout the valley, but especially in the agricultural areas to the east of the city. Cleaner water arriving from the hillslopes dilutes this contamination and improves water quality for rivers flowing through the city. We assume water treatment and sanitation only for the urban areas so under costs mapping we see water treatment and sanitation costs (per hectare) highest in the more densely populated west of the city. Note the need to upload improved urban area and population data. Under benefits mapping we see annual total water stress (fraction of demand not met non-zeros to be positive for much of the basin especially away from rivers in the city and in the populated areas to the south of the Shivapuri national park. Clearly in an urban context the realised domestic water stress for those with the economic resources to have piped water will depend on the pipeline distribution network and the location of water intakes and other sources. The number of people with the poorest quality drinking water non-zeros is zero throughout the city (because of water treatment) and the hillslopes (because of low contamination) but is highest in the populous agricultural areas near the city. Diarrhoea-relevant water quality is also lowest in populous areas around the city where sanitation is least. Total net erosion is best summarised by sub-catchments of order 5 using view by and shows generally low values in the valley but higher values on the steeper slopes draining in, especially in the south. Scenario for land use/cover change We first investigate the impacts of continued land use change in the catchment at rates similar to those in the last decade. We choose policy exercises and land use and cover change and then choose run land use change model. We set up the following scenario in which we reduce tree cover by 100% for all pixels deforested by continuing the current rate of change for 100 years with an index for fraction of forest degradation of 0.3 and allocating by accessibility to current roads only where land is not protected. Deforested areas are converted to cropland.
This scenario leads to a loss of tree cover of -15% for the catchment After running the simulation we see under results: maps that water balance decreases in some areas and increases in others. Using the inputs link we can see that the areas in which water balance increased were areas in which the loss in fog inputs on deforestation were much less than the loss in evapotranspiration output meaning more water available for runoff. Areas in which water balance declines are areas in which the loss in fog inputs on deforestation is greater than the loss in evapotranspiration output meaning less water available for runoff. Change in runoff non-zeros indicates small increases and decreases in the deforested areas cumulating to volumetric decreases downstream (especially in the south of the basin), though in percentage terms, these are small (<1 %) downstream. Since land use of the deforested areas was changed to cropland we also see changes in the human footprint on water quality non zeros with the footprint increasing significantly at the new croplands and also downstream of these, especially in the south where less of the catchment is protected. The change in land cover also leads to increases in annual total net hillslope soil erosion viewed by subcatchment of order 5, again especially in the south where most deforestation is expected to occur. Under costs mapping, as is to be expected water treatments costs (in % terms) non-zeros increase by upto 100% for all flows in the urban area (where water is treated). The statistics link indicates the average change for non-zero areas is 33%. Under benefits mapping the number of people (per hectare) with the poorest water quality non-zeros increases especially outside of the city (no water treatment) and in the south where conversion was greatest. What to do? Scenario for afforestation and protection A potential policy option to mitigate the impacts of recent and projected deforested is to implement and fund polciies to afforest the steepest slopes. We choose policy exercises and land use and cover change and then choose or define your own rule to set up sceanrio in which we increase per-pixel tree cover by 100% for areas in which slope gradient>=15 degrees except in urban areas and convert the land use back to natural.
This scenario leads to a recovery of 20% of baseline forest cover at the cost of 7.9% of cropland cover and 7.9% of pasture. This scenario leads to a change in water balance (non-zeros) that is positive in some places and negative in others, depending of the balance between gain in fog input and loss to evapotranspiration on afforestation. Afforestation on north facing slopes tends to increase water balance (less dominated by evapotranspiration) and on south facing slopes tends to decrease water balance. Change in runoff (non-zeros) indicates large volumetric increases of flow downstream (because of the enhanced fog inputs), though in percentage terms these are a fraction of a % in the major rivers. Since land use of the afforested areas was changed to natural we also see changes in the human footprint on water quality non zeros with the footprint decreasing significantly at the new forests and also downstream of these, though the downstream effects are small because of the significant inputs from valley agriculture and urban. The change in land cover also leads to reductions in annual total net hillslope soil erosion viewed by subcatchment of order 4 throughout. Under costs mapping, water treatments costs (in % terms) non-zeros decrease by upto 15% for all flows in the urban area (where water is treated). The statistics link indicates the average change for non-zero areas is -7.5%. Under benefits mapping the number of people (per hectare) with the poorest water quality non-zeros decreases but in few areas outside of the city. It appears much more difficult to improve water quality for those with the least than to further degrade it. |