Geoscientific Model Development | 2021

JULES-CN: a coupled terrestrial carbon–nitrogen scheme (JULES vn5.1)

 
 
 
 
 
 
 
 
 
 
 

Abstract


Abstract. Understanding future changes in the terrestrial carbon cycle is important\nfor reliable projections of climate change and impacts on ecosystems. It is\nwell known that nitrogen (N) could limit plants response to increased\natmospheric carbon dioxide and it is therefore important to include a\nrepresentation of the N cycle in Earth system models. Here we present the\nimplementation of the terrestrial nitrogen cycle in the Joint UK Land\nEnvironment Simulator (JULES) – the land surface scheme of the UK Earth\nSystem Model (UKESM). Two configurations are discussed – the first one\n(JULES-CN) has a bulk soil biogeochemical model and the second one is a\ndevelopment configuration that resolves the soil biogeochemistry with\ndepth (JULES-CN layer ). In JULES the nitrogen (N) cycle is based\non the existing carbon (C) cycle and represents all the key terrestrial N\nprocesses in a parsimonious way. Biological N fixation is dependent on net\nprimary productivity, and N deposition is specified as an external input.\nNitrogen leaves the vegetation and soil system via leaching and a bulk gas\nloss term. Nutrient limitation reduces carbon-use efficiency (CUE – ratio\nof net to gross primary productivity) and can slow soil decomposition. We\nshow that ecosystem level N limitation of net primary productivity\n(quantified in the model by the ratio of the potential amount of C that can\nbe allocated to growth and spreading of the vegetation compared with the\nactual amount achieved in its natural state) falls at the lower end of the\nobservational estimates in forests (approximately 1.0 in the model compared\nwith 1.01 to 1.38 in the observations). The model shows more N limitation\nin the tropical savanna and tundra biomes, consistent with the available\nobservations. Simulated C and N pools and fluxes are comparable to the\nlimited available observations and model-derived estimates. The\nintroduction of an N cycle improves the representation of interannual\nvariability of global net ecosystem exchange, which was more pronounced in\nthe C-cycle-only versions of JULES (JULES-C) than shown in estimates from\nthe Global Carbon Project. It also reduces the present-day CUE from a\nglobal mean value of 0.45 for JULES-C to 0.41 for JULES-CN and 0.40 for\nJULES-CN layer , all of which fall within the observational range.\nThe N cycle also alters the response of the C fluxes over the 20th\ncentury and limits the CO 2 fertilisation effect, such that the\nsimulated current-day land C sink is reduced by about 0.5\u2009Pg\u2009C\u2009yr −1 \ncompared to the version with no N limitation. JULES-CN layer \nadditionally improves the representation of soil biogeochemistry, including\nturnover times in the northern high latitudes. The inclusion of a\nprognostic land N scheme marks a step forward in functionality and realism\nfor the JULES and UKESM models.

Volume 14
Pages 2161-2186
DOI 10.5194/GMD-14-2161-2021
Language English
Journal Geoscientific Model Development

Full Text