I recently blogged on an excellent new paper by Steve Jennings and John Magrath on the changing nature of the seasons across a range of developing countries. One of the interesting side issues that emerged is that, while in most countries farmers’ perceptions fit the meteorological data, in a few others, farmers say that seasons, and particularly rainfall patterns have changed drastically since their parents’ time, but the meteorological data insist that nothing significant has occurred. What explains the disparity and who should we believe?
Some useful ideas come in a 2004 paper on Ethiopia by Elisabeth Meze-Hausken, published in the ‘Climate Research’ journal. She sets up the problem as follows:
‘Statistical analysis of rainfall chronologies was performed and contrasted with qualitative data collected through a survey and questionnaires. Fieldwork studies showed that local authorities, farmers and pastoralists perceived regional climate to have changed during the last few decades. Farmers explained that they have been changing their farming strategies by shifting to more drought-resistant crops as well as to a shorter agricultural calendar. They attributed this to a loss of the spring rains since ‘their father’s time’ (20–30 yr ago), as well as a shorter main summer wet period. However, rainfall measurements do not show a downward trend in rainfall.’
Meze-Hausken asked the question ‘If there is no proven change, has the need for rainfall changed during recent decades?’ For example increased demand for staple food (due to population growth) can be met by intensification or extensification of agriculture. If that increased pressure on the land leads to a crop failure, farmers may attribute it to drought (the traditional explanation) even if the rain fell as normal.
An alternative explanation is that other natural and human factors such as soil fertility, vegetation cover can reduce water availability, which farmers then interpret as due to declining rainfall, even without any actual measured change in rainfall itself.
Moreover, increased competition for land with the local pastoralists may have led to higher needs for stable and sufficient rainfall.
Steve Jennings, one of the authors of the Oxfam paper on seasonality, summarizes as follows:
‘The Ethiopia paper seems to say that farmer perceptions were of a compound variable (such as temperature + rainfall pattern + rainfall amount = agricultural drought) rather than the simple variable (rainfall amount) measured by weather stations. (No prizes for guessing he’s an environmental scientist by training……) This very much chimes with what we’ve supposed about the discrepancies between perceptions and data in Uganda (but sadly only had time to allude to in our paper). So the question then becomes why do people in some places seem to perceive changes in rainfall alone, whereas in others they perceive a compound variable. Is it what we ask them? Or is it that in places like Ethiopia, rainfall is so variable and the historical record so poor that it impossible for meteorologists to detect subtle changes that are obvious to communities?’
So who’s right, farmers or weathermen? Answer both, but they are measuring different things.
And if course it would be nice to have a similar exercise from the 1950s. Did 
tropical farmers back then say, ‘the weather hasn’t changed a bit since my granny’s time’, or did they too radiate nostalgia for the good old days? After all, it’s not just a British trait to complain about the weather.
