The Climate-Grain Production Relationship Quantified

Guest essay by David Archibald

There is now consensus that the Sun has now entered a quiet period. The first paper from the solar physics community predicting the current quiet period was Schatten and Tobiska’s 2003 paper “Solar Activity Heading for a Maunder Minimum?”. To date, Solar Cycle 24 has shown similar maximum SSN amplitudes to that of Solar Cycle 5, the first half of the Dalton Minimum:

Figure 1: Solar Cycle 24 relative to the Dalton Minimum

But what comes beyond that? Predicting the amplitude of Solar Cycle 24 was big business in the solar physics community with a total of 75 forecasts. There is only one forecast of the amplitude of Solar Cycle 25 to date. That forecast is Livingstone and Penn’s prediction of a maximum amplitude of seven. The first forecast, by Libby and Pandolfi, of the current quiet period is now over 40 years old. The fact that Libby and Pandolfi’s prediction got the detail of temperature changes to date right gives great credibility to it. Written in 1979, they forecast a warming trend for the rest of the 20^th century followed by a cold snap that might well last throughout the first half of the 21^st century. Specifically, Dr Libby is quoted by the Los Angeles Times as saying,

“we see a warming trend (by about a quarter of 1 degree Fahrenheit) globally to around the year 2000. And then it will get really cold – if we believe our projections. This has to be tested.” How cold? “Easily one or two degrees,” she replied, “and maybe even three or four degrees.”

The Libby and Pandolfi forecast was based on isotope ratios in tree rings and dates from a time before the corruption of tree ring science.

One commercial consequence of lower solar activity is that satellites will last longer in their orbits. Another is that agricultural production in the mid-latitudes will be affected. One of the most productive agricultural regions on the planet is the Corn Belt of the United States. Modern corn hybrids are tuned around maximizing the yield from the growing conditions experienced in the Corn Belt over the last 30 years with Growing Degree Days (GDD) to maturity ranging from 2200 to 2700. GDD is calculated from the day of planting by adding the maximum and minimum daily temperature in Fahrenheit, dividing by two and then subtracting 50 to produce the result. If the overnight minimum is less than 50°F, 50°F is used. The maximum is capped at 86°F as corn plants don’t grow any faster above that temperature. Daily temperature records for the Corn Belt start about 1900. The following graph shows the accumulation of GDDs for the periods 1901 – 1910 and 2001 – 2010 for Whitestown just northeast of Indianapolis in the southeast end of the Corn Belt:

Figure 2: Cumulative GDD for Whitestown, Indiana 1901 – 1910 and 2001 – 2010

The graph assumes a common planting date of 27^th April. The blue lines are the years 1901 – 1910 and the red lines are the years 2001 – 2010. They all stop on the date of first frost. Most of the growing seasons last decade had plenty of heat to get to maturity with up to 1,000 GDD in excess of the requirement at 2,500 GDD. A century before, the margin of safety was far less. Normal first frost for Whitestown is 10^th October. A century ago the earliest frost was five weeks before that on 3^rd September, 1908. Similarly, in the latter period the earliest date to get to 2,500 GDD was 15^th August. In the earlier period the last date to get to 2,500 GDD was almost six weeks later at 28^th September.

Farmers can adjust the type of crop they grow to suit their climatic expectations. Yield is directly proportional to GDD though as shown by the following graphic of corn and soybeans:

Figure 3: Yield relative to GDD (CHU) for Corn and Soybeans Source: Andy Bootsma, 2002: Potential Impacts of Climate Change on Eastern Canada

If a farmer plants a 2,200 GDD corn crop in the expectation of a cool or short season and the season turns out to have been capable of growing a 2,500 GDD, then he has foregone about 12% of the value of the later maturing variety. If he plants a 2,500 GDD variety and the season falls short though, most of the value of the crop will be lost. Wheat and barley require about 1,600 GDD and 1,400 GDD respectively. The highest wheat yield in Indiana in 2012 was 74 bushels/acre whereas the highest corn yield was 159 bushels/acre. Another factor in predicting grain output is the ability to switch to winter wheat in which a crop is planted in early September, germinates and then lies dormant under the snow blanket until the following spring.

A study in the 1980s of the effect of lower temperatures on Canadian wheat production found that a 1°C decrease would reduce the frost-free period by 15 days and that a 2°C decrease would not allow the crop to ripen before the first frost. Canadian wheat farmers have assured me though that they could switch to winter wheat and have a higher yield. In Manitoba, for example, the yield might be 71 bushels per acre for winter wheat compared to 51 bushels per acre for spring wheat. Growing winter wheat is riskier than spring wheat in that a hard frost before the first snow could kill the crop.

A further complication in trying to determine what the coming decline in temperature will do to grain production is that the area of the Corn Belt approximates to the region that was scraped flat by the Laurentide ice sheet. After the Wisconsin Glacier receded, the glaciated soils of the Midwest that are primarily north of Interstate 70 were covered with several feet of wind-blown loess deposits that came from the Great Plains that lie east of the Rockies. In Northern Illinois for example, in an area north of I-80, six to eight feet of loess deposits overlie glacier till. These soils are all primarily silt loam, silty clay loam, clay loam and clay. The water holding capacity of these soils are about 2 inches per foot. The counties in the Corn Belt with the highest productivity have deep fertile soils. Most of these soils were covered with prairie grass that over time raised the organic matter levels to between 2% and 5%. The resulting biological activity that developed in these soils made them very productive. These counties are also watered by natural rainfall that results from the Gulf of Mexico Pump. As the weather fronts move from west to east across the Rockies, we have the Great Plains that are mostly arid, but by the time the fronts reach eastern Nebraska, the moisture from the Gulf of Mexico is sucked north by the counter-clockwise flow of air that rotates around the low pressure fronts and drops the rain on the Midwest when it hits the cooler air from the north. Therefore the Corn Belt has the optimum combination of soil type, temperature and moisture. As growing conditions shift south, the soil types won’t be as good.

Friis-Christianson and Lassen theory enables us to predict temperature for a solar cycle if we know the length of the solar cycle preceding it. Thus Solheim et al have been able to predict that the average global temperature over Solar Cycle 24 will be 0.9°C lower than it was over Solar Cycle 23. Polar amplification also plays a part such that Svalbard, for example, in winter will experience a 6°C decline in temperature. Work on temperature records in the northeast United States suggest that the temperature decline in prospect for the Corn Belt is 2.0°C for Solar Cycle 24.

We can cross-check this expectation against modelled historic Total Solar Irradiance (TSI) data. Lean et al produced a reconstruction of TSI back to 1610. That is shown in Figure 4 following. Also shown is Livingstone and Penn’s prediction for Solar Cycle 25 amplitude converted to TSI by scaling against the Maunder Minimum. Shaviv in 2008 found empirically that a 1 watt/m² change in TSI was associated with (as opposed to cause directly) in a 0.6°C change in global average temperature. A fall in solar activity to levels reached in the Dalton Minimum, as per Lean’s data, would result in a decline of global temperature of 1.2°C, a little more than what Solheim’s group is projecting. Solar Cycle 4, the cycle preceding the Dalton Minimum, was 13.6 years long, about a year longer than Solar Cycle 23. Libby and Pandolfi’s prediction of a temperature decline of up to 4°F translates to 2.2°C. Through TSI, this would require a fall of 3.7 watts/m² which is greater than the range in Lean’s modelled data for the period since 1610. This may mean that Libby and Pandolfi are correct and Lean’s model needs adjusting.

Figure 4: Projecting the decline in Total Solar Irradiance

Working through the effect on GDDs, a return to TSI conditions of the Dalton Minimum can be expected to reduce US corn production by perhaps 20% to 25%. This equates to the increase in corn production over the last ten years from mandated ethanol. US grain and soybean production of about 500 million tonnes per annum is sufficient to feed 1.2 billion vegetarians. The amine profile of wheat can be approximated by a diet of 70% corn and 30% soybeans, otherwise those things are fed to animals at about a 25% protein conversion efficiency. Corn and soybeans would be the diet of involuntary vegetarianism. The rest of the world does not have the luxury of US agriculture’s latent productivity.

Figure 5: US Corn and Wheat Prices 1784 to 2013

Figure 5 shows the effect of the low temperatures of the Dalton Minimum on corn and wheat prices in the United States. The absolute peak was associated with the eruption of Mt Tambora. Also evident is the period of high and volatile prices associated with the cold temperatures of the mid-19^th century.

Figure 6: Major wheat exporting countries

A return to the climatic conditions of the Dalton Minimum is likely to take Russia, Kazakhstan and the European Union out of the export market. The other countries will have some reduction in wheat available for export. Colder is also drier and thus a number of major grain producers such as India and China, currently largely self-sufficient, will experience shortfalls from their requirements.

Figure 6: Imports and exports of grain by continent

Figure 6 above shows net exports of grain by continent with the Arab countries as a separate region. Those countries are the biggest grain importing block on the planet. Soybeans are not included in this graphic. China has become the major soybean importer at 60 million tonnes per annum. In terms of protein content, that equates to about 180 million tonnes of wheat per annum. The Chinese convert those soybeans to animal protein in the form of pig meat.

Countries in the Middle East North Africa (MENA) region have been in the news recently. Further detail on their import dependency is shown in Figure 6 following.

Figure 6: MENA region domestic and imported grain by country

In Figure 6, the population size of each country is shown by the size of the bar. The blue component of the bar shows how much of each country’s grain requirement is grown domestically and the red component denotes the imported share. Countries are shown from west to east as per the map. A proportion of the Egyptian population already suffers from malnutrition. A current wheat prices, it costs about $1 per day to keep someone fed in terms of bulk grain. The oil exporting countries in the graphic can afford to feed their populations, with some countries feeding others as well. Saudi Arabia has been keeping Yemen above water and more recently took on Egypt too.

Figure 7: An animal model of population growth and collapse

All the countries of the MENA region have seen their populations grow to well in excess of their inherent carrying capacity. A combination of deteriorating climate and ongoing world population growth can be reasonably expected to cause a spike in grain prices to levels last seen in the 19^th century. It is also possible that sufficient grain may not be available at any price in some regions. Populations models from the animal kingdom provide some guidance as to how events might unfold. A good example is the snowshoe hare and lynx of North America. The snowshoe hare population collapses to less than 10% of its peak on a roughly ten year cycle, followed by the lynx. Taking the example of Egypt, the current population is twice the level that can be supported by its grain production. If the food supply to that country falls below the minimum required to maintain public order, then the distribution system for diesel and fertiliser will break down and domestic grain production would also be affected.

The starving populations of Egyptian cities will fan out into the countryside and consume whatever they can chew which will include the seed grain. That will ensure that domestic grain production will collapse. The population of Egypt might fall to 10% of its carrying capacity which would be 5% of its current level. Any starvation in the MENA region is likely to trigger panic buying by other governments in the region and beyond with consequent effects on established trade patterns.

UPDATE:

The Excel spreadsheet for the Whitestown data used in this essay is here Whitestown-all-years (.xlsx file)