Monitoring the depth of the atmospheric boundary layer by GPS radio occultation signals

Transcription

1 Monitoring the depth of the atmospheric boundary layer by GPS radio occultation signals S. Sokolovskiy, D. Lenschow, Z. Zeng, C. Rocken, W. Schreiner, D. Hunt, Y.-H. Kuo. R. Anthes University Corporation for Atmospheric Research Boulder, CO 2010 AGU Fall Meeting, San Francisco, CA, December 13-17

2 Outline: Effects of the atmospheric boundary layer (ABL) on radio occultation (RO) signals Different definitions of the ABL top from RO Comparisons of the variabilities of the ABL depth derived from bending angle and water vapor from COSMIC RO data Structural uncertainty of the determination of the ABL depth from RO

3 ABL is the layer directly affected by the Earth's surface 1) Turbulently mixed layer underlying stably stratified free atmospehere 2) Separated from the free atmosphere by the interfacial layer. In many cases the interfacial layer is sharp, characterized by temperature inversion and strong decrease of humidity N 77.6 P T P T w 2 The decrease of humidity on top of ABL results in decrease of refractivity and strong effect on propagation of RO signals. This has been used for determination of the ABL depth from RO

4 Effects of the ABL on propagation of RO signals Strong increase of bending angle (BA) at the ABL top (when pronounced) + fading of amplitude due to defocusing Fluctuation at lower heights due to (i) multipath propagation, (ii) turbulence The effects (i) and (ii) can be separated by transforming RO signals from coordinate to impact parameter representation z incident wave inversion turbulence stably stratified atmosphere layer N equivalent phase screen

5 Studies of the ABL depth from GPS RO Loss of signal by a GPS RO receiver (PLL) due to defocusing (von Engeln et al., 2005) Bending angle lapse (Sokolovskiy et al., 2007; Ratnam and Basha, 2010) Specific humdity gradient (Ao et al., 2008) Refractivity break point (Guo et al., 2010) Estimation of the depth of convective layers from fluctuation of the transformed (FSI) amplitude (Sokolovskiy et al., 2007)

6 Determining the height of ABL (or strongest inversion layer) Maximum bending angle lapse Maximum lapse of water vapor pressure retrieved from refractivity by 1Dvar Both methods can be used when the ABL top is well pronounced

7 Transform of RO signals from coordinate to impact parameter representation - for calculation of the bending angles. Amplitude of the transformed signal is not sensitive to horizontally homogeneous N structures and it is sensitive to turbulence. Case A Case B

8 Transform to impact parameter representation allows vertical localization of turbulent layers max. lapse of LSW max. LSW Fluctuation of the amplitude of the transformed RO signal can be characterized by scintillation index (S4) or by local spectral width (LSW)

9 When the ABL top is pronounced, the strongest turbulence is observed near the top (Wyngaard and LeMone, 1980) The height of max. BA lapse (max. water vapor gradient) is correlated with the height of max. LSW and the height of max. lapse of LSW This establishes a statistical relation between the main physical definition of the ABL top (capping turbulence) and the definition based on WV or BA lapse (more convenient for practical use).

10 ABL depth (from BA lapse) from COSMIC RO (only profiles penetrating below 0.3 km) D J F M A M J J A S O N

11 Distribution of the ABL depth (max. BA lapse > 1E-2 rad) over North America Summer: most sharp inversion layers (pronounced ABL top) over the ocean and plains; less over mountains Winter: - fewer strong inversion layers over continent, more over the ocean southwards - shallower ABL over continent - deeper ABL over the ocean

12 Study of the variability of the ABL depth (by BA and WV) from COSMIC RO data, boxes - regions selected for the study

13 Interannual variation of the ABL depth (land) max. BA lapse max. WVP lapse

14 Diurnal variation of the ABL depth (land) max. BA lapse max. WV lapse

15 Interannual variation of the ABL depth (oceans) max. BA lapse max. WV lapse

16 Diurnal variation of the ABL depth (oceans) max. BA lapse max. WV lapse

17 Structural uncertainty BA - from bending angle lapse; WV - from water vapor lapse (1) - all data; (2) - BA lapse > 0.01 rad; (3) add. min. ret. alt. < 0.3 km Small BA lapse may not necessarily correspond to the ABL top. May introduce bias in averaging. Possible constraints: - magnitude of the BA or WV lapse? - ABL depth? - percentile of lapse, ratio first/second maxima? (Ao et al. 2008) - median instead of mean? (no tunable parameters) ?

18 Structural uncertainty On the other side, the deeper ABL (the smaller BA lapse) the larger is the variability? Interannual variation Diurnal variation BA - from bending angle lapse; WV - from water vapor lapse (1) - all data; (2) - BA lapse > 0.01 rad, rad (~50 percentile)

19 When max. BA lapse is larger than some value, the averaged LSW above the layer decreases, i.e. the inversion layer is capping turbulence. This is consistent with the main physical definition of the ABL top. This may provide a justification for use of this value (may depend on region) as a constraint in the statistical estimates of variations of the ABL depth.

20 Summary Determination of the ABL top from RO based on strong gradient of humidity (BA, N, etc.) is consistent with the main physical determination based on turbulence. Any of the RO-retrieved profiles (bending angle, refractivity, humidity) can be used for monitoring of the ABL depth. Existing RO data allow studies of the geographical, seasonal, interannual and diurnal variations (diurnal variations over the oceans has never been observed before). For the weather variations, higher density of RO soundings is needed. For statistical studies, it is important to consider the structural uncertainty. Averaging by constraining on sharpness of the ABL top results in lower estimates of the mean ABL depth and the magnitude of its variations.