SEARCH

SEARCH BY CITATION

Keywords:

  • agriculture;
  • carbon;
  • corn;
  • light-use efficiency;
  • net primary production;
  • wheat

Abstract

Remote sensing of net primary production (NPP) is a critical tool for assessing spatial and temporal patterns of carbon exchange between the atmosphere and biosphere. However, satellite estimates suffer from a lack of large-scale field data needed for validation, as well as the need to parameterize plant light-use efficiencies (LUEs). In this study, we estimated cropland NPP with the Carnegie-Ames-Stanford-Approach (CASA), a biogeochemical model driven by satellite observations, and then compared these results with field estimates based on harvest data from United States Department of Agriculture National Agriculture Statistics Service (NASS) county statistics. Observed interannual variations in NPP over a 17-year period were well modelled by CASA, with exceptions mainly due to occasional difficulties in estimating NPP from harvest yields. The role of environmental stressors in agriculture was investigated by running CASA with and without temperature and moisture down-regulators, which are used in the model to simulate climate impacts on plant LUE. In most cases, correlations with NASS data were highest with modelled stresses, while the opposite was true for irrigated and temperature resistant crops. Analysis of the spatial variability in computed LUE revealed significantly higher values for corn than for other crops, suggesting a simple parameterization of LUE for future studies based on the fraction of area with corn. Absolute values of LUE were much lower than those reported in field trials, due to uncommonly high yields in most field trials, as well as overestimates of absorbed radiation in CASA attributed to bias from temporal compositing of satellite data. Total NPP for US croplands, excluding Alaska and Hawaii, was estimated as 0.62 Pg C year−1, representing ∼20% of total US NPP, and exhibited a positive trend of 3.7 Tg C year−2. These results have several implications for large-scale carbon cycle research that are discussed, and are especially relevant for studies of the role of agriculture in the global carbon balance.