%0 Figure %A Lantz C Baldos, Uris %A W Hertel, Thomas %D 2013 %T Global results of the evaluation experiments (1961–2006) %U https://iop.figshare.com/articles/figure/_Global_results_of_the_evaluation_experiments_1961_2006_/1011569 %R 10.6084/m9.figshare.1011569.v1 %2 https://iop.figshare.com/ndownloader/files/1479394 %K crop yields %K time series estimates %K supply response %K climate change impacts %K capita consumption %K land use change %K food demand response %K cropland use %K crop production %K crop price %K simple %K model %K Environmental Science %X

Figure 4. Global results of the evaluation experiments (1961–2006). The panels show the per cent changes in crop production (a), yield (b), land use (c) and price (d) computed from actual data (dashed line) and from the simulation results (colored bars). The experiments consists of (E1) exogenous per capita consumption, (E2) fixed food demand response, (E3) short to medium run supply response, (E4) restricted intensive margin, (E5) targeted crop yields, (E6.a) combined exogenous per capita consumption, restricted intensive margin and targeted crop yields, and (E6.b) combined fixed food demand response and short to medium run supply response.

Abstract

Global agricultural models are becoming indispensable in the debate over climate change impacts and mitigation policies. Therefore, it is becoming increasingly important to validate these models and identify critical areas for improvement. In this letter, we illustrate both the opportunities and the challenges in undertaking such model validation, using the SIMPLE model of global agriculture. We look back at the long run historical period 1961–2006 and, using a few key historical drivers—population, incomes and total factor productivity—we find that SIMPLE is able to accurately reproduce historical changes in cropland use, crop price, crop production and average crop yields at the global scale. Equally important is our investigation into how the specific assumptions embedded in many agricultural models will likely influence these results. We find that those global models which are largely biophysical—thereby ignoring the price responsiveness of demand and supply—are likely to understate changes in crop production, while failing to capture the changes in cropland use and crop price. Likewise, global models which incorporate economic responses, but do so based on limited time series estimates of these responses, are likely to understate land use change and overstate price changes.

%I IOP Publishing