Assessing the impact of measurement time interval when calculating wind speed means and trends under the stilling phenomenon

In a globally warming climate, a weakening in observed near-surface wind speed has been recently termed as 'global stilling', showing a worldwide average trend of -0.140ms(-1) dec(-1). The precise cause(s) of the land-surface stilling remains uncertain and led to this first assessment of how the two most common measurement time intervals of daily mean data averaged wind speed being either: (1) four-synoptic times (0000, 0700, 1300 and 1800 UTC; WS) or (2) 24-h wind run measurements (WR) can affect the estimation of wind speed averages and trends. This was performed across Spain for 1961-2011 (12 stations) and 1979-2008 (19 stations), where WS andWR daily wind speed observations were simultaneously recorded. Results indicate that mean wind speed is 0.24ms(-1) statistically greater for WS than WR measurements annually, being seasonally dependent with major differences in July (0.49ms(-1)) and minor in December (-0.01ms(-1)); that WS (-0.057ms(-1) dec(-1)) shows a more negative wind speed trend when compared with WR (-0.011ms(-1) dec(-1)) annually (and seasonally) for 1979-2008, but few trend differences are statistically significant; and that the percentage of stations showing stilling is greater for WS(63.2%) than WR(36.8%) annually (and seasonally) for the shortest period. In contrast, differences are almost negligible for 1961-2011. These findings may have direct implications for interdisciplinary areas such as agriculture and hydrology and wind renewable energy, and highlight the need of improving our understanding on the causes associated with wind speed declines under a climate change scenario.