Weather Variability and Macro-Agronomy

A few days ago we talked about weather and Micro-Agronomy, now we are going to look at a bigger scale. 

Besides knowing how local weather affects their own yields, growers also want information that will help them optimally market their crop. For this we need to know how weather is affecting crops across the entire Corn-Belt, not just in our backyards. Is there predictive weather information that can help? Is it possible to correlate historic temperature or precipitation data with crop prices in a way that would help us know, perhaps a few weeks ahead, if crop prices are likely to go up or down? Of course we know that there are other factors besides weather that impact crop prices, such as foreign production, international demand, support prices, just to name a few. However, weather is the predominant factor affecting total US grain production which drives crop prices. So, it would therefore be of interest for us to know how seasonal weather variations might affect total corn or soybean production. Much controversy has been made of late regarding changes in the earth’s ambient surface temperature, so let’s start there. USDA yield data and weather data from the Midwest Regional Climate Center (MRCC) over the last 50 years or so have provided some insights. Here is what the data tells us.

1. Irrespective of the annual temperature variations across the Midwest, temperature has little bearing on the aggregate harvest numbers for corn or soybean yields. That is, whether our winters, springs, or summers have been warmer or cooler than average, the change in ambient temperature has to be quite significant to affect a change in total U.S. harvest numbers. In general, it seems that while temps above or below average may favor some sections of the Midwest, it is also likely to hinder crop production in other areas. The result is that total U.S. corn and soybean harvest numbers, and hence crop prices, have tended to be insensitive to the variations in temperature that typically occur each year. (Charts below)

Out of the last 50 years only two exceptions are noted. The first was in 1974 when cold temperatures in April and May resulted in late planting, and an early frost that fall curtailed some crop development before maturity. The other exception was in 1983.

In that year yields were significantly reduced across the corn belt as a result of July and August temperatures that averaged 4 degrees F above normal, and pollination and seed fill were compromised. Although rainfall and soil moisture was generally adequate throughout the Corn-Belt that summer, high day and night-time temperatures and low humidity contributed to yield losses by accelerating plant respiration.  

2. While precipitation often does impact aggregate harvest numbers for major crops in the Midwest, there is no direct correlation between the average prevailing temperature and precipitation. That is, yearly or monthly temperature fluctuations have had no bearing on whether the corn-belt region gets more or less precipitation than average during any month or year. While it might seem reasonable to expect warmer weather to facilitate the development of storm systems that bring rain into the Midwest, this does not appear to be the case. As an aside, the months of the year with the greatest temperature variations are December through March, which can vary in average temperature by as much as 25 degrees. The months with the least temperature variations are June through September, which have varied less than 10 degrees Fahrenheit in average temperature over the last 50 years.

3. As for precipitation, U.S. corn and soybean harvest numbers tend to be resilient to swings in the amount of precipitation received annually. This is not to say that precipitation amounts don’t have an effect on crop yields and prices, only that efforts to use aggregate annual or monthly precipitation amounts for the Midwest to predict U.S. crop harvest numbers is a complex task. The second two charts show the trend lines of seasonal precipitation (May through August rains) and U.S. corn and soybean harvest numbers for the years of 1960 through 2013. Significant reductions in corn and soybean yields are noted for the years 1974, 1983, 1988, 1993, and 2012, only two of which, 1988 and 2012, are correlated with significant reductions in seasonal precipitation. In 1988, the precipitation deficit was in May and June, and in 2012, the deficit occurred in July and August. As mentioned above, yield reductions in 1983 were a result of exceptionally warm temperatures, not a lack of precipitation, in July and August. The low harvest numbers from 1993 were due to a very wet spring and summer and flooding throughout much of the Midwest. The lower harvest numbers for 1974 were due to cool spring temps and late planting and an early frost throughout much of the corn belt.

Predicting crop pricing by trying to pin down the effects of precipitation or temperature on harvest numbers is difficult. Only five exceptions seem to have occurred in the past 50 years. Of those five exceptions, two were a result of drought (1988 and 2012), one was a result of too much rain in season (1993), one was a result of temperatures being too warm during the season (1983), and one resulted from temperatures being too cold in the spring with a follow-up frost that came early in the fall (1974).