Comparison of nitrogen inputs on forest obtained by deposition modelling and observed soil nitrogen accumulation

Forests capture more reactive nitrogen gases and dust particles than short vegetation. However, this effect averages out in large scale deposition models with a low spatial resolution, due to the limited proportion of forests in grid cells. Here we illustrate the need of using high resolution input data to accurate capture nitrogen (N) deposition inputs in fragmented forested areas, such as common in the Netherlands. We applied the Dutch atmospheric transport model OPS on 127 forest stands across the Netherlands in combination with data on the fate of atmospheric N deposition, i.e. observed soil N r accumulation rates and estimates for forest N uptake and N leaching to water during the period 1990-2023 using different spatial resolutions for the OPS input data on land use and surface roughness. Results showed that the calculated average deposition from 1990 to 2023 on the 127 forest stands was approximately 34 kg N ha -1 yr -1 when using a standard spatial resolution of 1 km x 1 km. Increasing the calculation resolution to 250 m x 250 m gives an average deposition of approximately 38 kg N ha -1 yr -1 while using data for forests alone, which corresponds to the land use for the measuring locations, an average deposition of 44-45 kg N ha -1 yr -1 was calculated. The estimated mean N r deposition based on observations of soil N r accumulation and estimates of forest N uptake and N leaching was 51 kg N ha -1 yr -1 . Despite the uncertainties involved in the estimates of both mean N deposition and its fate (standard errors varying between 10-20%), results indicate that even the use of relatively high spatial resolution grid cells, such as 250 m x 250 m, is too coarse to capture the enhanced N deposition on forests in regions with fragmented forested areas.