Before we get into detail about plant biology and ecology, let us see what is being said about the effect of CO2 emissions on climate. Specifically, it is said that the earth warmed during the twentieth century. In truth, taking the earth as a whole, including parts of the earth where no-one lives and where crops are not grown (i.e most of the earth), a very small amount of warming occurred, about half a degree overall, none of which can be attributed to CO2 emissions in any robust scientific way – it has all the hallmarks of natural variation. Some parts of the earth cooled and some warmed.
We are told that the effects of future warming, coupled with decreasing rainfall, will cause terrible difficulties for crop production in the tropics and equatorial areas, because many crops in those regions are already on the limits for heat damage, and are already badly stressed due to drought (we will see the evidence in later posts that the best solution for crops that are heat or drought stressed is increased atmospheric CO2). We are told that the likely future scenarios will be worst for countries in the tropics and equatorial zones. Seldom are we told that the same scenarios predict much-needed warmth to northern Canada, northern Europe and Siberia, bringing vast tracts of land into agricultural use, which with the increased atmospheric CO2 will enable greatly increased food production – and that in countries that have the infrastructure, the technology, and the capital to make the best use of it.
But is it true that trends are pointing towards impending catastrophe in sub-Saharan Africa and South Africa? Not at all, if trends from the nineteenth century are anything to go by.
Firstly, look at the chart of temperature changes on land across the world during the twentieth century. Note: data records not available, or incomplete, over the full 100 years in the grey regions.
We see that the Indian sub-continent has experienced a little warming (none in some parts); southern China has cooled, as has central Africa and central South America. Mid-USA regions (the great plains) have remained stable or cooled slightly. Those northern and southern latitudes that can benefit from warmer weather have seen a welcome rise in temperature. South-west Africa (Namibia), which is mostly desert anyway, has warmed. All in all, a very acceptable picture.
What about precipitation? See below for the change through the 20th Century.
West Africa has seen a noticeable decrease. Sub-Saharan Africa a very small decrease. Central and southern Africa, slight increase. Southern South America, Australia, USA and Canada, and northern Europe and Siberia show an increase. If we split this out by latitude to see the trends over the last century we have the following:
This looks very favourable. Between the tropics, rainfall has been pretty steady over 100 years. For the more temperate zones the precipitation has been increasing.
With regard to temperature, a world map of temperature changes over 100 years was presented above, but we now turn to records year-by-year from specific monitoring stations to get the detailed picture. One of the difficulties here is finding continuous measurements made at the same location for over 100 years, but they do exist, though they are a relatively small subset.
So are there are unfavourable long-term warming trends? Let us look at South Africa, which is cited as the region likely to see the most drastic drop in agricultural yield due to heat stress, according to the global warming hypothesis (as reiterated in National Geographic for June 2009, which suggests a 47% decrease in corn production by 2030). Surely we would expect to see some evidence of this in South African climate trends. In fact, we do not. To their credit, even the South African government departments admit that there is no statistical significance in temperature trends in their country that would suggest a warming pattern.
Firstly, Cape Town:
If we rigged and cherry picked the data to start from the 1960s we might infer a trend, but we see that over the longer period of nearly 130 years there is just no statistical support for a long-term warming trend, and the warmest period was the 1930s.
At Durban the trend over 120 years is cooling, but again if we look only from the 1960s we could falsely infer warming. In the 20th Century the warmest period was the 1930s.
No warming trend over more than 120 years at Port Elizabeth, but a short record starting from the 1960s could be used to infer a warming trend. Again the 1930s were the warmest.
Central South Africa at Kimberley. No noticeable warming trend over 110 years except at the start of the 20th Century, leading up to the 1930s.
What about Zimbabwe to the north?
Bulawayo in Zimbabwe shows no warming trend over 120 years unless one selectively looks only from the 1970s.
What about Madagascar to the east?
Again, a very slight warming trend could be proposed from the 1970s, but the record over 120 years shows that this is nothing of significance: the 1920s were by far the warmest.
Going north-west to sub-Saharan Africa:
Timbuktu in Mali has an incomplete dataset, but records extend for over 100 years, and no warming trend is discernible (unless one deliberately looks only from the 1970s to 2000).
Across north-west to Egypt:
No overall warming at Aswan, but a dataset starting in the 1960s would give a false picture. As usual, the 1930s are the warmest.
East across to India:
No evidence of warming over 130 years (look at the scale on the temperature axis as well – very stable!).
South-east to Australia:
Central Australia has no statistically significant warming trend, looking over 130 years, but a shorter dataset from the late 1970s would give a false impression.
Across to South America:
Brazil has no discernable warming trend.
South to Paraguay:
Once again, no warming trend over 120 years. The 1940s and 1950s are the hottest.
We finish with the USA, starting in Texas:
A definite cooling trend over 120 years in Texas.
Across to North Carolina:
No warming trend noted in Salisbury, NC, when viewed over 130 years. It would be easy to make a warming case by cherry picking the data only from 1980 to present, though. The 1930s-40s are particularly warm.
Across to Nebraska, the great corn belt:
No warming trend over 120 years. The 1930s are the warmest.
To the north of the great plains and the mid-West:
No trend over 130 years. Warming could be inferred only by starting the dataset from the 1960s. The 1930s again stand out as the warmest.
It should be clear that the global warming lobby have their work cut out to convince us that the habitable regions of the earth suitable for growing crops have experienced warming trends in the last 130 years of any magnitude that are dangerous for crops. Parts of the world that have experienced noticeable warming (Canada and northern Europe and Russia) are now in a better position to grow crops, and constitute a vast area. It is abundantly clear that many parts of the earth cooled from the 1930s and 1940s to the 1960s and 1970s, and a warming in the 1980s and 1990s was merely a recovery to the long-term average, or within natural variation.