Is there a coherence in observed and projected changes in riverine low flow indices across Central Europe?

Central Europe, a region composed of Austria, Czech Republic, Germany, Hungary, Poland, Slovakia and Slovenia, occupying over 1 million km2 and inhabited by >150 million people, has recently experienced several extensive and severe hydrological droughts (e.g., 2015, 2018, 2019) that affected a wide range of sectors. These drought events were broadly attributed to climate change, but often without, or with only limited, scientific assessment. There is a common belief that warmer climate and more frequent heat waves are likely to lead to more severe hydrological droughts in the future. However, there is lack of robust findings that would hold at either national or regional scale, so that comparison of these findings is of broad relevance and interest in the region. This systematic-style review attempts to identify the evidence for: (1) historical trends in observed data and (2) changes in model-based projections for the future, in low river flow and hydrological drought indices in rivers of Central Europe. In the context of this review, focusing on directions rather than magnitude of historical and future changes, we were treating hydrological droughts and low flows as synonyms. To address these questions, we searched Web of Science and Scopus databases and screened 976 abstracts to identify 68 articles fulfilling all inclusion criteria from which metadata were extracted and analyzed. The results show that overall, trends detected in observation records have more frequently downward (i.e. meaning decreased low flows or increased drought hazard) than upward direction (53% vs. 11%). However, the frequency of evidence reporting decreases in future low flows is lower for future projections than for historical trends (43% vs. 53%), and even more convincingly, nearly three times more evidence items point out at upward trends in the future (31% vs. 11%). This shows that there is a low coherence between observed and projected indices of low river flows in Central Europe. Catchment topography appears to be an important factor affecting trend direction: in mountainous catchments only 33% of evidence items pointed out at increasing hydrological drought hazard, whereas in other types of catchments this frequency was almost doubled. A striking difference in dominant future direction of changes between studies based on SRES and RCP scenarios was identified: in the former only 6% of evidence items pointed out at less severe hydrological drought hazard in the future, while in the latter it was 42%. Finally, systematic review approach enabled us to identify some knowledge gaps, such as studies on low flow historical trends in lowland and upland catchments as well as in large river basins.