It’s routinely claimed that climate-change-induced drought in Syria was a major factor in triggering the Syrian civil war, the Syrian refugee crisis and the rise of ISIS. But are these claims supported by the data? This post investigates this question.
We begin with this quote from a Climate and Security article, which claims that the recent drought in Syria was the worst in Syria’s long history and is not alone in doing so:
From 2006-2011, up to 60% of Syria’s land experienced, in the terms of one expert, “the worst long-term drought and most severe set of crop failures since agricultural civilizations began in the Fertile Crescent many millennia ago” …… This has led to a massive exodus of farmers, herders and agriculturally-dependent rural families from the countryside to the cities. Last January, it was reported that crop failures ….. just in the farming villages around the city of Aleppo, had led 200,000 rural villagers to leave for the cities.
Now look at Figure 1, which shows the GHCNv2 annual rainfall record for Aleppo (data from KNMI Climate Explorer). Average annual rainfall during the 2006-2011 period was only 9% lower than average annual rainfall over the preceding 55 years. The driest year during the period (2011) was only the seventh driest on record and 2006-2011 was only the 13th driest six-year period on record. Clearly the crop failures in the farming villages around Aleppo – which undoubtedly occurred – weren’t caused by a drought of Biblical proportions. In fact there doesn’t seem to have been a drought at Aleppo at all:
Figure 1: GHCN v2 monthly rainfall record, Aleppo. The graphic is as it appears on Climate Explorer except for the added shading
But Aleppo is a single data point in a large country. What happened elsewhere?
First we will look at the rainfall data from other stations in Syria to see if there were rainfall deficiencies that could have contributed to crop failures in other areas. Figure 2 shows the locations of the seven stations in Syria with GHCNv2 rainfall records going back to the 1950s relative to the distribution of cropland. They give fairly good country-wide coverage:
Figure 2: Stations with rainfall records in cropland areas
Figure 3 shows the annual rainfall records for the seven stations with the “drought years” highlighted as before. Average rainfall over 2006-2011 was below the pre-2006 average at four stations (Deir Ezzor -31%, Palmyra -22%, Lattakia -17% and Aleppo -9% as already discussed ) but above it at three (Kamishli +3%, Damascus +5% and Hama +15%). The average for all seven stations was 7% below the pre-2006 average, decreasing to 4% when only the five “cropland” stations (Lattakia, Aleppo, Kamishi, Hama and Damascus) are considered:
Figure 3: GHCN v2 rainfall records for Kamishli, Lattakia, Hama, Damascus, 2006-2011 period highlighted in yellow
Decreases of only 4-7% in average annual rainfall clearly don’t qualify as drought conditions, but there is of course more to drought than just rainfall. To get a true indication of drought intensity we must look at the Palmer drought severity index (PDSI), which takes other factors such as soil moisture content and temperature into account, and a paper by Al Riffai et al (2012) provides annual PDSI estimates between 1960 and 2009 for each of the five “agroecolocial” zones shown in Figure 4:
Figure 4: Agroeconomic zones in Syria, Figure 1 of Al Riffai et al, 2012
Comparing this map with Figure 2 shows that crops are grown only in and around Zones 1, 2, 3 and 4 (5 is desert), so if drought caused widespread crop failures it must have affected these four zones. Figure 5 shows Al Riffai et al’s self-calibrating PDSI index for the four zones between 1960 and 2009, with the 2006-2009 period highlighted in yellow. Note that negative values indicate drier conditions:
Figure 5: Drought Index values, agroeconomic zones 1, 2, 3 and 4, modified from Figure 3 of Al Riffai et al, 2012
The post-2006 drought is visible only as abrupt downward spikes in 2008 in Zones 3 and 4 and as a weaker downward spike in the same year in Zone 1. But none of these spikes, reaches the minus 4 “extreme” PDSI drought threshold and none of them are the lowest values since 1960.