Dry deposition of reactive nitrogen to different ecosystems across eastern China: A comparison of three community...

The assessment of nitrogen ecosystem loads mostly use the method of sampling observation combined with numerical model to estimate the spatial distribution pattern of nitrogen dry deposition flux. The selection of models is important which directly affects the reliability of the deposition flux results. In this study, the performance of three widely used models (WRF-Chem, EMEP, CMAQ) are compared. The dry deposition fluxes of typical active nitrogen components over eastern China showed uncertainties by a factor of 0.5 ~ 2 between the oxidized nitrogen (OXN) results of the three models and the observation network while the reduced nitrogen (RDN) simulation results are underestimated by a quarter of the observation reports. These three models show different results on four typical ecosystems: simulation of EMEP got the highest for OXN dry deposition flux on each ecosystem (urban 14.94 ± 4.92kgN ⋅ ha^-1 ⋅ yr^-1, cropland/grassland 5.53 ± 5.11kgN ⋅ ha^-1 ⋅ yr^-1, forest 4.75 ± 4.32kgN ⋅ ha^-1 ⋅ yr^-1, water bodies 1.48 ± 1.53kgN ⋅ ha^-1 ⋅ yr^-1); WRF-Chem has the highest value of RDN on the urban (8.91 ± 6.44kgN ⋅ ha^-1 ⋅ yr^-1) and water bodies (1.01 ± 1.44kgN ⋅ ha^-1 ⋅ yr^-1) while EMEP is highest in cropland/grassland (3.42 ± 3.43kgN ⋅ ha^-1 ⋅ yr^-1) and forest (2.34 ± 1.94kgN ⋅ ha^-1 ⋅ yr^-1). CMAQ is in medium range for both OXN and RDN simulations on each ecosystem. Compare with the critical loads, CMAQ generates more exceeded critical load areas than WRF-Chem and EMEP on cropland/grassland and forests ecosystem. For water bodies, WRF-chem and CMAQ showed higher exceeding critical load areas than EMEP. In summary, EMEP generally underestimates while the CMAQ and WRF-Chem model would overestimate the impacts on the ecosystems. So, policy implementation needs special attention accounting the difference of simulation effect with different models.