Publication Abstracts

Glotter et al. 2016

Glotter, M.J., A.C. Ruane, E.J. Moyer, and J.W. Elliott, 2016: Evaluating the reliability of reanalysis as a substitute for observational data in large-scale agricultural assessments. J. Appl. Meteorol. Climatol., 55, no. 3, 579-594, doi:10.1175/JAMC-D-15-0120.1.

Future projections of food security require historical agricultural assessments to validate and improve crop yield models. Insufficient observational climate networks often force historical assessments to rely on reanalyses- numerical weather prediction models assimilated with observational data - for inputs to crop models. However, reanalyses are subject to some degree of unavoidable error, since many climate variables (including precipitation and solar radiation)are not directly assimilated. It is therefore important to assess whether these issues may be sufficient to compromise the validation exercise. In this study we test the reliability of reanalysis output for simulating maize yields in the U.S., where observational data are extensive. We drive a highly parallelized version of the Decision Support System for Agrotechnology Transfer (DSSAT) crop model with climate inputs from a combination of data sources: originaland bias-corrected reanalyses (CFSR and AgCFSR, respectively), and observation-based precipitation and solar radiation. We find that yield estimates are more robust when driving DSSAT with observation-based precipitation than with reanalysis precipitation, although errors in observation-based climate inputs can themselves also affect modeled yields. Bias-corrected reanalysis produces the most reliable estimates. We note however that modeled outputs appear oversensitive to interannual climate fluctuations, suggesting that even perfect weather inputs may not accurately reproduce observed yields. We find relatively little sensitivity of yields to daily-scale fluctuations, suggesting that hard-to-collect and unreliable daily records may not be necessary for bias correction. Agricultural assessments may benefit most from a hybrid combination of low-frequency monitoring networks and high-resolution models.

Export citation: [ BibTeX ] [ RIS ]

BibTeX Citation

@article{gl04000e,
  author={Glotter, M. J. and Ruane, A. C. and Moyer, E. J. and Elliott, J. W.},
  title={Evaluating the reliability of reanalysis as a substitute for observational data in large-scale agricultural assessments},
  year={2016},
  journal={J. Appl. Meteorol. Climatol.},
  volume={55},
  number={3},
  pages={579--594},
  doi={10.1175/JAMC-D-15-0120.1},
}

[ Close ]

RIS Citation

TY  - JOUR
ID  - gl04000e
AU  - Glotter, M. J.
AU  - Ruane, A. C.
AU  - Moyer, E. J.
AU  - Elliott, J. W.
PY  - 2016
TI  - Evaluating the reliability of reanalysis as a substitute for observational data in large-scale agricultural assessments
JA  - J. Appl. Meteorol. Climatol.
VL  - 55
IS  - 3
SP  - 579
EP  - 594
DO  - 10.1175/JAMC-D-15-0120.1
ER  -

[ Close ]

➤ Return to 2016 Publications

➤ Return to Publications Homepage