Mapping Above Ground Carbon Storage and Sequestration in Thoria Watershed, India: A Spatially Explicit Ecosystem Service Assessment Using InVEST Model

Globally, the commons provide important ecosystem functions and services and contribute towards sustainable development and human population. About 2.5 billion people depend on these commons for many material and non-material benefits and to support and improve their livelihoods. However, the contribution of the commons to the smallholder farmers and the economy are not very clear. Quantitative analysis of the contribution of ecosystem services will help to develop policies to better support their management and allocate resources to protect them. In this study, we assess the carbon storage and sequestration of the commons in the Thoria watershed, India using the Integrated Valuation of Ecosystem Services and Trade-offs (InVEST) model. InVEST is a suite of spatially explicit models for evaluating trade-offs associated with land-use induced ecosystem service changes, developed by the partners of the Natural Capital Project. The InVEST Carbon Storage and Sequestration model uses maps of land use along with stocks in carbon pools to estimate the amount of carbon currently stored in a landscape or the amount of carbon sequestered over time. Aboveground biomass comprises all living plant material above the soil (e.g., bark, trunks, branches, leaves). The model maps carbon storage densities to LCLU rasters which may include classes such as forest, pasture, or agricultural land. The carbon stocking contents of over a hundred samples are measured in the field among major land cover types. The average of carbon stocking values is used and assigned to designated land cover types. The LCL U map is developed by using a machine learning algorithm with Landsat 7 imageries and ground truth points collected by the local collaborators. The model summarizes results into raster outputs of storage, sequestration, and value, as well as aggregate totals. The ground truth data and remotely sensed data are used to calibrate the parameters of the model inputs. We also perform two scenario analyses: 1) expansion of forest, 2) expansion of agricultural land. The two land-use scenarios are generated using proximity-based approaches where the user determines which land-use type can be converted to, as well as the type of patterns based on proximity of the edge of focal land use. Results of the scenario analysis demonstrate the potential changes of carbon stocking to land cover and land-use change.