Snowpack Variability and Change across the Middle East and North Africa with a View to Hydrologic Impacts

Mid-latitude snowpacks are an important dynamic component of regional climates and, often, they form a key societal resource as well, delivering water during the melt season that can be used immediately or stored for future use. While hemisphere-wide snow trends and some regional snow patterns have been carefully studied in recent years, a number of key regional systems have received less attention. An example is the Middle East, which is a region historically sensitive to climate variability and change, and which contains snowpacks that have been shown to be important inputs to regional water resources, as well as, influencing regional climate patterns.

Focusing on the Middle East (and the smaller snowpacks of northwestern Africa), this presentation aims to (i) quantify each year’s snowpack development and recession over recent decades, and (ii) understand historical and present-day interannual to decadal variability and change in the snowpacks, with specific reference to timing in the annual cycle, and potential implications for water resources (as well as other societal impacts, such as flash flooding). The presentation draws on a range of data sources, including satellite-based products, station data, and model reanalyses. Variation in the spatial fields of snow are considered, as well as variability summarized in indices, averaging regional snow conditions for (i) Northwestern Iran and Southern Caucasus (NWIC), including the Zagros Mountains and (ii) Eastern Turkey (ETKY) including the Taurus Mountains. In addition, smaller-scale indices quantify snow variations in (i) the mountains of Lebanon and (ii) the Atlas Mountains of northwestern Africa.

For satellite-based snow products, the Interactive Multisensor Snow and Ice Mapping System (IMS) is a daily snow cover product that is unique in its combination of moderate length (1999-present) and relatively high resolution (24km). It is a professional best estimate analysis derived primarily from visible satellite products, but over time, has included additional satellite products and other data sources. Analyzing 1999-2016, it is found that for both NWIC and ETKY, the peak snow cover extent typically occurs in late January, and usually becomes near-zero by late June. The decay rate is initially somewhat faster for NWIC. The mean late-January extent for NWIC is slightly less than that for ETKY (275,000km² v 325,000km²), however, the interannual variation for both domains is very large, with both regions ranging from about 450,000km² to 125,000km² (a slightly higher range is found for NWIC). By early April, mean extent declines to about 50,000km² for NWIC and 100,000km² for ETKY. In early April, interannual variability is still large for NWIC (ranging 130,000-15,000km²) and especially for ETKY (ranging 250,000-15,000km²). These variations and the implied melt patterns can be expected to have significant implications for local and regional water resources, including drainage into the Tigris-Euphrates river system.

For station data, the Global Historical Climatology Network – Daily (GHCND) is drawn upon for observations of snow depth, precipitation and temperature. Systems are developed to analyze station snow depth for a target day in the annual cycle, including drawing on temperature and precipitation data to help inform on days when snow depth is not reported, and especially to help inform on days that have zero snow depth. Networks of approximately 20 stations are identified with sufficient data for both ETKY and NWIC, and are analyzed from 1981 to present. The IMS and GHCND indices are compared at the 5-day (pentad) timescale, as well as monthly/seasonal timescale, over 1999-2016. Agreement of indices using IMS and GHCND is very high.

Analysis of IMS and GHCND snow trends is undertaken, motivated by the body of literature documenting the decline in recent decades of snow cover extent (SCE) averaged over the Northern Hemisphere in spring, when the snow albedo feedback is considered to be most effective due to prevailing radiation and snow cover conditions. This general pattern of downward trend in SCE is shown to be present in the Middle East in the IMS SCE data. Analysis of GHCND trends is ongoing, but at least for analyses currently well advanced (indices of ETKY) there is good agreement between IMS and GHCND on patterns of trend in relation to annual cycle. Analysis of related atmospheric circulation, temperature and precipitation changes is also ongoing, to further inform the possible causes of the SCE trends.

The interannual variability in SCE and depth variations is diagnosed and interpreted in terms of regional atmospheric circulation patterns as revealed by reanalysis datasets. On the interannual timescale, the center-of-axis of atmospheric circulation influence is located somewhat differently for the NWIC index compared to that found for the ETKY index, which is related to blocking activity further upstream near 0-5^oE longitude, permitting arctic cold air outbreaks in conjunction with cyclonic circulation over Turkey. Indeed, the heavy snowfall in early January 2017, which included the ETKY domain as well as extending unusually westward into the Balkan Peninsula, is shown to have a circulation signature consistent with teleconnection structures diagnosed using 1999-2016 ETKY IMS indices in January.

Analyses for northwestern Africa and Lebanon SCE in relation to atmospheric circulation demonstrate how the higher resolution IMS product at 4km (available since 2004) permits diagnosis of the spatial details of interannual SCE variation associated with atmospheric circulation and local mountains. These results especially illustrate the potential for developing downscaling systems using the IMS data as target information. It is planned to further develop these analyses with data from other satellite-based and reanalysis-based products. The overall assessments reported here and ongoing, are considered to form a basis for underpinning risk assessments of snow-related societal impacts today, as well as providing a baseline for assessments of possible future change in these regions.