Thursday, April 16, 2015

UTD 200S: Assessment of LiDAR measurements against sonic anemometry

The UTD team is working on the intercomparison between the radial velocity measured by the UTD WindCube 200S and the 3D velocity measurements carried out with the CSAT3 sonic anemometers installed over the met-tower. In this post measurements performed in proximity of the sonic anemometer installed on the South-East boom at height of 50 m are shown. For this test the UTD 200S measured with an elevation angle of 11.49 deg, and the azimuthal angle between the laser beam and the SE boom was 98 deg. The three wind velocity components measured by the CSAT3 are retrieved in order to obtain the equivalent radial velocity seen by the UTD 200S. In Fig. 1 the comparison between the radial velocity retrieved from the sonic data and the one measured by the UTD 200S clearly shows a very good agreement between the two measurement techniques. These data were acquired from 2015-03-22 16:00:08.806244 UTC till 2015-03-22 16:14:38.285902 UTC. For this analysis no temporal correction was performed, assessing the time synchronization of the LiDAR carried out with the embedded GPS. 

Fig. 1: Comparison between the 3D velocity data obtained from the CSAT3 sonic anemometer and the radial velocity measured by the UTD 200S.
  In the following some numbers to quantify the LiDAR accuracy:

Mean value for the lidar radial velocity: -3.07 m/s
Mean value for the sonic radial velocity : -2.91 m/s
Standard deviation for the lidar radial velocity: 0.32 m/s
Standard deviation for the sonic radial velocity: 0.42 m/s 

The higher standard deviation of the sonic data is connected to the higher sampling rate (20 Hz for the sonic and 2 Hz for the LiDAR) and smaller sampling volume (30 cm for the sonic, 50 m for the LiDAR).
In Fig. 2 a zoom in of the comparison between the sonic and LiDAR data is reported.
Fig. 2: Comparison between the radial velocity measured by the UTD 200S and the one retrieved by the sonic anemometer CSAT3 installed on the SE boom at 50 m height.
Finally, accuracy of the LiDAR measurements can be better appreciated when LiDAR data are compared with the wind shear measured by the sonic anemometers deployed at different heights.
Fig. 3: Intercomparison between the UTD 200S measurements performed in proximity of the sonic anemometer at 50 m height and the wind data acquired by the sonic anemometers deployed at different heights.

The regression analysis of the LiDAR data vs the sonic anemometer data shows a slope of  1.072 and an R-squared value of 0.6682 (Fig. 4).

Fig. 4: Linear regression between LiDAR data and sonic anemometer data.

2 comments:

  1. Thanks, Valerio - this looks interesting. When the 200S data and the v_radial_sonic are plotted against each other, what's the slope and the R^2?

    ReplyDelete
  2. Hi Julie, thanks for your comment. I added another figure for the linear regression between lidar and sonic data. The slope is 1.072 and the R-squared value is 0.6682. This values are obtained from 15 minute data. However, there is a significant variability already within the same measurement.

    ReplyDelete