Wednesday, August 15, 2012
The USDA has published an early estimate of what they think this year's United States average corn yield will be - a much depressed 123.4 bushels/acre due to the current record-breaking heat and drought in the US. Since they also maintain statistics for average yields back to 1866 (!!), I made the graph above showing the context for this year's estimate (assuming it turns out to be reasonably accurate).
You can see that prior to the late 1930s, yields were pretty much flat with fluctuations year-to-year. Then began a long and roughly linear rise in yields due to improving agricultural technology. Year-to-year weather causes fluctuations around the trend, and the latest fluctuation is particularly large (though not off the scale of what has been seen before).
Hitherto, I've been of the view that there was no climate change signal visible in these yields, but the latest data point is perhaps starting to raise a small question mark there.
Recall that the United States generally got wetter during the 1960s to 1990s, likely due to natural sea surface temperature fluctuations, but has been getting drier since then due to some (arguable) combination of reversal of the natural fluctuation and overall climate drying due to excess carbon dioxide:
So, if we were going to see a climate related yield signal, we'd expect to see it mainly since the late 1990s. With the latest data point, there is now no trend in the yields if we just look at data back to 2002:
If you go back earlier, then there is still a positive upward trend. I doubt that this is a statistically significant trend change (I'd have to work harder than I have time for right now to tell for sure). But it's definitely something to watch.