3.3.2 Meteorological conditions during extreme events across Cumbria
Figures 6 and 7 (and Figures Supp. 1 & 2) show time series of meteorological variables during the 4 extreme Cumbrian rainfall events that occurred since 2005. Each event was associated with a frontal system, the 2005, 2009 and 2015 events being classified as atmospheric rivers, where enhanced horizontal water vapour transport from the Atlantic Ocean occurs (Lavers et al ., 2011, 2013; Matthews, Murphy, McCarthy, Broderick, & Wilby, 2018). The meteorological time series show that, at many of the locations, during the main periods of rainfall (indicated by the grey shaded area in Figures 6b, 7b, Supp. 1b & 2b) relatively low RH (in the context of the results presented above) occurs along with high wind speeds (Figures 6a, 7a, Supp. 1a & 2a) and an increase in air temperature (Figures 6, 7c, Supp. 1a & 2a). A notable exception to these general patterns are the differences inRH between sites for the 2009 event where Keswick and Walney were at approximately 90-95% RH during the main rainfall period but the Shap and Walney sites were at, or close to, saturation. This was also the case at the higher elevation site of Great Dun Fell where the wind speeds were very high but the RH remained at 100% throughout the entire period of rainfall (Figures 7a and 7b respectively). Similarly, but for the 2015 flood event, in a clearing in Gisburn Forest, Lancashire the RH remained very high or at saturation for a large part of the storm (Figure 7b). There is also consistency in that Keswick, which is a less exposed site at relatively low elevation on the lee side of one of the highest regions of mountains, tends to have lower wind speeds, higher temperatures and lower RH than the other locations during all extreme events considered (Figures 6, 7 and Supp. 1 & 2). It is worth highlighting the caveat that the Keswick, Shap and Warcop sites are located on the leeward side (relative to the dominant south westerly flows) of mountain ridges such that the observed favourable conditions for Ewc loss shown are not necessarily representative across the region as a whole: as shown by the less favourable conditions for Ewc loss at Gisburn Forest and Great Dun Fell. These results show that even during the 4 most extreme events in Cumbria over the last 15 years, high windspeeds and surprisingly low RH provide favourable conditions for significant Ewc loss at some locations. Consequently the need to estimate Ewc losses across large catchments, particularly in complex mountainous terrain, requires a representation of the spatial variability of meteorological controls which will be challenging where only sparse meteorological observations are available.