Assessing the sensitivity of urban ecosystem service maps to input spatial data resolution and method choice

Ecosystem service (ES) mapping frequently uses secondary data and value-transfer methods to map services over broad extents at coarse resolutions, possibly causing poor prediction accuracy. Although ES map quality has received some recent attention in the literature, little is known about the accuracy of these maps in urban contexts or about the factors that influence this accuracy. To address this issue, we quantitatively compared and validated ES maps in a heterogeneous urban landscape to generate insight into ES map accuracy in these environments. Using aboveground biomass carbon storage as an example, we examined how input data resolution and assessment method affect the accuracy of urban ES maps. Two mapping methods were employed: (1) maps based on ecosystem components involved in carbon storage (trees and lawns) and (2) maps based on land-cover proxies and data at coarse and fine spatial resolutions. We compared carbon storage predicted by these methods to that estimated by using field-collected data to examine the accuracy of predictions and spatial variation therein. Different methods and data produced similar study area-wide estimates; however, the spatial distribution of estimates varied among methods. Estimates using ecosystem components agreed with the actual observations better than the proxy-based estimates, although map accuracy was improved by using higher resolution land-cover data. Thus, when study area-wide estimates suffice for decision making, proxy-based methods and coarse-resolution data should provide adequate assessments. Detailed ecosystem structure and composition data are needed when fine-resolution, spatially-explicit estimates are required.