Evaluating parameterisations of subgridscale variability with satellite data

Transcription

1 Evaluating parameterisations of subgridscale variability with satellite data Johannes Quaas Institute for Meteorology University of Leipzig Acknowledgements Vera Schemann and Verena Grützun (Max Planck Institute for Meteorology, Hamburg) Torsten Weber (Climate Service Centre Hamburg)

2 Contents The Tompkins cloud scheme in the ECHAM5 GCM Evaluation of the moments of the total water path distribution Critical relative humidity as a simple metric for variability Evaluation with supersite observational data? Scale dependency of total-water variance 2/38

3 Prognostic subgrid-scale PDF of total water mixing ratio PDF PDF of total water mixing ratio follows a β-function prognostic equations for variance and skewness Total water model already includes equations for water vapour, cloud liquid- and ice water mixing ratio symmetric or positively skewed distributions Tompkins, J. Atmos. Sci. 22; Lohmann and Roeckner, Clim. Dyn /38

4 Total water Total water PDF PDF PDF PDF Prognostic subgrid-scale PDF of total water mixing ratio Total water Deep convection Turbulence generates skewness dissipates variance via detrainment Total water Precipitation reduces skewness Tompkins, J. Atmos. Sci. 22 4/38

5 Evaluation of total water path variability Spatial PDF of vertically integrated total water path (TWP) (sum of precipitable water (spatially interpolated), liquid water path and ice water path) MODIS resolution: 5x5 km² PDF at GCM resolution ~2x2 km² MODIS TWP in Grid-Box S S S 15 W 14 W 13 W 3 [kg m-2] Weber, Quaas, Räisänen, QJRMS 211 5/38

6 Evaluation of total water path variability Spatial PDF of vertically integrated total water path (TWP) (sum of precipitable water (spatially interpolated), liquid water path and ice water path) MODIS resolution: 5x5 km² PDF at GCM resolution ~2x2 km² MODIS TWP in Grid-Box S MODIS TWP statistics MODIS TWP statistics cloud cover:.48 cloud cover:.48 mean kg m skewness: mean. kg m-2 variance: 1.21 kg m values: 296 skewness:.35 variance: 1.21 kg-2m-4 values: S 15 W 14 W 13 W 3 [kg m-2] Probability Density 18 S PDF [kg m ] m-2 34 kg -2 Weber, Quaas, Räisänen, QJRMS 211 6/38

7 Evaluation of total water path variability Randomly picked examples from individual grid-points Diagnose skewness and variance from any shape Weber, Quaas, Räisänen, QJRMS 211 7/38

8 Mean value of total water path [kg m-2] MODIS satellite data F DP Evaluation of total water path variability Model Satellite data Weber, Quaas, Räisänen, QJRMS 211 8/38

9 Variance of total water path [kg2 m-4] MODIS satellite data MODIS satellite data F DP Evaluation of total water path variability Model Satellite data Model Satellite data Weber, Quaas, Räisänen, QJRMS 211 9/38

10 Skewness of total water path MODIS satellite data F DP Evaluation of total water path variability Model Satellite data Weber, Quaas, Räisänen, QJRMS 211 1/38

11 Critical relative humidity as a simple metric Cloud cover parameterisation rc =.7 f cloud fraction r relative humidity rc critical humidity Sundqvist et al., Mon. Wea. Rev /38

12 Critical relative humidity as a simple metric f cloud fraction q specific humidity qt total humidity qs saturation Le Treut and Li, Clim. Dyn /38

13 Critical relative humidity as a simple metric f cloud fraction q specific humidity qt total humidity qs saturation For the choice Δq = γ qs and assuming saturation within the cloud, this is equivalent to the critical relative humidity scheme with rc = 1 - γ Le Treut and Li, Clim. Dyn /38

14 Critical relative humidity as a simple metric rc =.7 f cloud fraction r relative humidity rc critical humidity rc small subgrid-scale variability γ large Sundqvist et al., Mon. Wea. Rev /38

15 Global annual mean profile Sundqvist et al. (MWR, 1978) cloud parameterisation rc =.7 in free troposphere rc =.9 at surface Quaas, J. Geophys. Res., /38

16 Critical relative humidity as a simple metric rc =.7 f cloud fraction r relative humidity rc critical humidity gridbox mean observable! (for < f < 1) Quaas, J. Geophys. Res., /38

17 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover ECMWF re-analysis (ERA) GCM-oriented CALIPSO cloud product (GOCCP) T63 grid (1.8 x1.8 ) daily data 27 Quaas, J. Geophys. Res., /38

18 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation Supersaturated ice clouds? ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover ECMWF re-analysis (ERA) GCM-oriented CALIPSO cloud product (GOCCP) T63 grid daily data 27 Quaas, J. Geophys. Res., /38

19 ERA-Interim/GOCCP critical relative humidity: annual-mean distribution large subgrid variability small variability Quaas, J. Geophys. Res., /38

20 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover AIRS satellite data (A acending orbit/daytime) (D descending/night) Atmospheric InfraRed Sounder (Aqua) AIRX3STD T63 grid (1.8 x1.8 ) daily data for 23 Quaas, J. Geophys. Res., 212 2/38

21 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover AIRS satellite data (A acending orbit/daytime) (D descending/night) Tompkins (JAS, 22) cloud parameterisation rc diagnostic from f and r Quaas, J. Geophys. Res., /38

22 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover AIRS satellite data (A acending orbit/daytime) (D descending/night) Tompkins (JAS, 22) cloud parameterisation Tiedtke (MWR, 1993) cloud parameterisation rc diagnostic from f and r Quaas, J. Geophys. Res., /38

23 Global annual mean profile Sundqvist et al. (MWR, 1978) parameterisation ERA-Interim/CALIPSO ERA: relative humidity CALIPSO: cloud cover AIRS satellite data (A acending orbit/daytime) (D descending/night) Fits of Sundqvist to data Tompkins (JAS, 22) cloud parameterisation Tiedtke (MWR, 1993) cloud parameterisation Quaas, J. Geophys. Res., /38

24 Subgrid scale variability from lidar measurements? Strategy Differential absorption lidar (DIAL) Hamburg (H. Linné) 3 g/kg height (m) time Grützun, Quaas, Ament, in preparation 24/38

25 Subgrid scale variability from lidar measurements? Strategy Differential absorption lidar (DIAL) Hamburg (H. Linné) 3 g/kg height (m) h time 5 km 3 m 2 Model Grützun, Quaas, Ament, in preparation 25/38

26 Subgrid scale variability from lidar measurements? Strategy Differential absorption lidar (DIAL) Hamburg (H. Linné) 3 g/kg height (m) h time 5 t (-1km) -1km t +1km 3 m km wind t (+1km) 2 Model t (km) Grützun, Quaas, Ament, in preparation 26/38

27 Subgrid scale variability from lidar measurements? COSMO model as virtual reality High-resolution model (2.8 km) temporal sampling at one point virtual lidar spatial sampling at one timestep virtual GCM gridbox Grützun, Quaas, Ament, in preparation 27/38

28 Subgrid scale variability from lidar measurements? mean [g kg-1] Temporal = virtual lidar Spatial = virtual GCM grid-box 6 6 Total water mean 3 12(h) time Time (h) m 25 m 12 m 5 m 3 time (h)12 24 Time (h) 28/38

29 Subgrid scale variability from lidar measurements? mean [g kg-1] Temporal = virtual lidar 6 6 Total water mean 3 skewness Spatial = virtual GCM grid-box 12(h) time 24 Time (h) 3 Total water skewness m 25 m 12 m 5 m time (h) Time (h) 3-3 Grützun, Quaas, Ament, in preparation 29/38

30 Total water variance spectrum [kg2 kg-2 m-1 ] Scale dependency of total water variance kmkm 1 km 1 km 1 m 1 m wavenumber [m-1 ] Schemann, Stevens, Grützun, Quaas, J. Atmos. Sci., submitted 3/38

31 Total water variance spectrum [kg2 kg-2 m-1 ] Scale dependency of total water variance kmkm 1 km 1 km 1 m 1 m wavenumber [m-1 ] Schemann, Stevens, Grützun, Quaas, J. Atmos. Sci., submitted 31/38

32 Scaled total water variance spectrum General circulation model (Tropics) Regional NWP model LES 19 km 1 km 1 m Schemann, Stevens, Grützun, Quaas, J. Atmos. Sci., submitted 32/38

33 Scaled total water variance spectrum General circulation model (Tropics) Regional NWP model LES 19 km 1 m 1 cm Schemann, Stevens, Grützun, Quaas, J. Atmos. Sci., submitted 33/38

34 Scale dependency of total water variance T63 resolution (~19 km) T127 resolution (~1 km) Variance of total water mixing ratio [g2 kg-2 ] Resolved Subgrid Tompkins parameterisation Subgrid extrapolated using slope Subgrid extrapolated using -2.3 slope.6 Schemann, Stevens, Grützun, Quaas, J. Atmos. Sci., submitted 34/38

35 Conclusions Spatially high-resolved satellite data allow to evaluate total water path variance allows for useful conclusions too little variance in Tompkins scheme, need for negative skewness Critical relative humidity is a metric available from satellite data including vertical resolution problematic for ice clouds, dependent on assumptions It is difficult to use supersite measurements as a reference for higher moments Total water mixing ratio variance scaling follows a power-law with an exponent of about -2. allows to evaluate and improve parameterisations 35/38

36 Conclusions Spatially high-resolved satellite data allow to evaluate total water path variance allows for useful conclusions too little variance in Tompkins scheme, need for negative skewness Critical relative humidity is a metric available from satellite data including vertical resolution problematic for ice clouds, dependent on assumptions It is difficult to use supersite measurements as a reference for higher moments Total water mixing ratio variance scaling follows a power-law with an exponent of about -2. allows to evaluate and improve parameterisations 36/38

37 Conclusions Spatially high-resolved satellite data allow to evaluate total water path variance allows for useful conclusions too little variance in Tompkins scheme, need for negative skewness Critical relative humidity is a metric available from satellite data including vertical resolution problematic for ice clouds, dependent on assumptions It is difficult to use supersite measurements as a reference for higher moments Total water mixing ratio variance scaling follows a power-law with an exponent of about -2. allows to evaluate and improve parameterisations 37/38

38 Conclusions Spatially high-resolved satellite data allow to evaluate total water path variance allows for useful conclusions too little variance in Tompkins scheme, need for negative skewness Critical relative humidity is a metric available from satellite data including vertical resolution problematic for ice clouds, dependent on assumptions It is difficult to use supersite measurements as a reference for higher moments Total water mixing ratio variance scaling follows a power-law with an exponent of about -2. allows to evaluate and improve parameterisations 38/38