What do we know about past changes in the water cycle of Central Asian headwaters? A review

• We provide a comprehensive review of more than 100 studies and data sets.
• We discuss the underlying data sets, the methods, and the related uncertainties.
• Knowledge on the magnitude and the spatio-temporal pattern of change is incomplete.
• Major research gaps concern changes in seasonal snow cover and river runoff.
• There is a need for detailed change attribution studies.

We have reviewed about 100 studies on past changes in climate, snow cover, glaciers and runoff in Central Asian headwater catchments, which have been published in the past 20 years. We included studies published by Central Asian researchers in Russian language, which are usually not easily accessible to international researchers.Most studies agreed on general warming trends in Central Asia with acceleration since the 1970s, but varied with regard to seasonal changes and the magnitude of the warming. Most studies also confirmed that glaciers in the Tien Shan and the Pamir continue to retreat and to shrink, though only little is known about mass and volume changes. Only few studies investigated changes in seasonal snow cover, and they suggested a decrease in maximum snow depth and a reduction in snow cover duration. The studies on runoff trends in the high mountain areas of Central Asia indicated a complex response of catchments to changes in climate. It appears that catchments with a higher fraction of glacierized area showed mainly increasing runoff trends in the past, while river basins with less or no glacierization exhibited large variations in the observed runoff changes.We conclude that our knowledge is still incomplete in particular with regard to the magnitude and the spatio-temporal patterns of changes in the water cycle of Central Asian headwater catchments. The limitations in our knowledge are due to (1) the scarcity of reliable and appropriate data sets especially for the glacio-nival zone; (2) methodological limitations of trend analysis; (3) the heterogeneity in both spatial and temporal extent of the available analyses, hampering the synthesis to a regional picture; and (4) the insufficiently understood interactions between changes in highly-variable climate parameters, the cryosphere, and the hydrological response of individual headwater catchments.Finally, there is a need for sound attribution studies linking the observed hydrological changes in individual catchments to particular processes triggered by climatic and cryospheric changes. This research gap needs urgently to be closed as projections of future hydrological changes are of vital importance for water management in Central Asia.