Modeling of net ecosystem exchange and its components for a humid grassland ecosystem

Delphine Lawton, Paul Leahy, Ger Kiely, Kenneth A. Byrne, Pierluigi Calanca

Research output: Contribution to journalArticlepeer-review

Abstract

We measured the net ecosystem exchange (NEE) of a managed humid grassland in southwest Ireland from 2002 to 2004 with an eddy covariance (EC) system. In addition, a process-based biogeochemical model (PaSim) incorporating land management practices such as grazing and grass harvesting was used to simulate the carbon dynamics. The modeled NEE of 2.6, 2.7 and 3.4 t C ha-1 (uptake) for 2002, 2003 and 2004 compares well with the measured NEE of 1.9, 2.6 and 2.9 t C. ha-1. There is good agreement between the model output and the EC observations in the growing season but not so good in the winter period. The year-on-year increase in measured NEE is partly attributed to a circa 4% year-on-year increase in annual photosynthetic photon flux density (QPPFD). The year of lowest NEE (2002) was associated with highest rainfall (1785 mm) and lowest QPPFD. In the wettest year, grass harvesting was delayed by a month, resulting in a reduced NEE. The management of grassland in regions of high rainfall is dependent on weather conditions. If wet conditions become more prevalent (e.g., as a result of climate change), grasslands in such regions may shift from intensive to extensive management with further reductions in NEE. The reasonable agreement between the model predictions and the EC measurements demonstrates the potential of the model for applications such as upscaling EC measurements to regional scales and predicting responses of grasslands to climate change.

Original languageEnglish
Article numberG04013
JournalJournal of Geophysical Research: Biogeosciences
Volume111
Issue number4
DOIs
Publication statusPublished - 28 Dec 2006
Externally publishedYes

Fingerprint

Dive into the research topics of 'Modeling of net ecosystem exchange and its components for a humid grassland ecosystem'. Together they form a unique fingerprint.

Cite this