Observed spatial variability of boundary-layer turbulence over flat, heterogeneous terrain

Abstract

In spring 2013, extensive measurements with multiple Doppler lidar systems were performed. The instruments were arranged in a triangle with edge lengths of about 3 km in a moderately flat, agriculturally used terrain. For six mostly cloud-free convective days, vertical velocity variance profiles were compared for the three locations. On the aver-5 age over all considered cases, differences between variances at different sites were about three times higher than between those derived from measurements by different lidars at the same site. For all investigated averaging periods between 10 min and 4 h, the differences were not significant on the average when considering the statistical error. However, statistically significant spatial differences were found in several individual 10 cases. These could not be explained by the existing surface heterogeneity.In some cases, nearby energy balance stations provided surface fluxes that were not suitable for scaling the variance profiles. Weighted-averaged values proved to be more applicable, but even then, the scaled profiles showed a large scatter for each location. Therefore, it must be assumed that the intensity of turbulence is not always 15 well-determined by the local heat supply at the Earth's surface. Instead, a certain dependency of turbulence characteristics on mean wind speed and direction was found: thermals were detected that travelled from one site to the other with the mean wind when the travel time was shorter than the large-eddy turnover time. At the same time, no thermals passed for more than two hours at a third site that was located perpendic-20 ular to the mean wind direction in relation to the first two sites. Subsidence prevailing in the surroundings of thermals advected with the mean wind can thus partly explain significant spatial variance differences existing for several hours.

Evaluation of large-eddy simulations forced with mesoscale model output for a multi-week period during a measurement campaign

“…Figure 9 shows the vertical velocity, potential temperature and mixing ratio variances for 24 April and 5 May (11:00-12:00 UTC). For 24 April, the lidar-based variances (solid purple lines) of the vertical velocity, the actual temperature and the absolute humidity were each recently published by Maurer et al (2016), Behrendt et al (2015) and Muppa et al (2016). They also provide data for the cumulustopped boundary layer of 5 May; these are analyzed for the first time in the present paper.…”

“…1b). From these measurements spatially representative values of the surface fluxes are derived and provided by Maurer et al (2016). A weighted average (w.av.)…”

“…A weighted average (w.av.) of the five stations with the fraction of the respective land-use class in an area of 30 km × 30 km centered around the KITcube site is calculated (see Maurer et al, 2016, for further details). The weighted average is marked by purple stars in Fig.…”

“…The variances of vertical velocity w 2 , potential temperature θ 2 and mixing ratio q v 2 from PALM and UCLA-LES are compared to variance profiles from lidars located at the KITcube site for a 1 h period (11:00-12:00 UTC) for the clear-sky situation of 24 April and the shallow-cumulus situation of 5 May. Three different lidars, namely the Doppler lidar WindTracer WTX combined with the Doppler lidar WLS7 (Maurer et al, 2016) from KIT, the rotational Raman lidar (RRL; Hammann et al (2015); Behrendt et al (2015)) and the water vapor differential absorption lidar (WVDIAL; Muppa et al (2016)) from the University of Hohenheim were operated simultaneously during IOPs of HOPE, allow us to compare different lidar- based higher-order moments with the LES to discuss the turbulence structure of the boundary layer on these 2 days. Figure 9 shows the vertical velocity, potential temperature and mixing ratio variances for 24 April and 5 May (11:00-12:00 UTC).…”

“…The vertical velocity variance from Doppler lidar exhibits a maximum at around 0.5z i and shows a second smaller peak around 0.9z i . A longer averaging period of about 3 h would lead to a decrease in the height of the lower maximum to about 0.3z i (Maurer et al, 2016), which emphasizes that the chosen averaging time might be too small to receive robust w 2 statistics comparable to LES. This is confirmed by the large differences between the given virtual measurements.…”