Structure of observable meteorological state variables during - - PowerPoint PPT Presentation

Structure of observable meteorological state variables during transitions in the stably stratified nocturnal boundary layer

Carsten Abraham & Adam H. Monahan

University of Victoria

June 14, 2018

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 1 / 14

Why investigate transitions in the SBL?

Inaccurate representations of turbulence in weather and climate models under stable conditions cause misrepresentations in the near-surface temperature and wind profiles:

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 2 / 14

Why investigate transitions in the SBL?

Inaccurate representations of turbulence in weather and climate models under stable conditions cause misrepresentations in the near-surface temperature and wind profiles:

◮ forecast of fog, frost, or dew formation

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 2 / 14

Why investigate transitions in the SBL?

Inaccurate representations of turbulence in weather and climate models under stable conditions cause misrepresentations in the near-surface temperature and wind profiles:

◮ forecast of fog, frost, or dew formation ◮ harvesting wind energy

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 2 / 14

Why investigate transitions in the SBL?

Inaccurate representations of turbulence in weather and climate models under stable conditions cause misrepresentations in the near-surface temperature and wind profiles:

◮ forecast of fog, frost, or dew formation ◮ harvesting wind energy ◮ assessment of pollutant dispersal and air quality

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 2 / 14

Why investigate transitions in the SBL?

Inaccurate representations of turbulence in weather and climate models under stable conditions cause misrepresentations in the near-surface temperature and wind profiles:

◮ forecast of fog, frost, or dew formation ◮ harvesting wind energy ◮ assessment of pollutant dispersal and air quality

What can we learn about the physics of the transitions in the SBL from observational data?

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 2 / 14

Two regime behaviour at Cabauw, Netherlands

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 3 / 14

The hidden Markov model (HMM)

◮ estimate from observable state variables an unobservable Markov chain ◮ statistical model to systematically detect regime behaviour when state variables

in state space overlap

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 4 / 14

HMM analysis of two regime behaviour at Cabauw

;

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 5 / 14

Frequency of transitions

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

wSBL to vSBL transition

Boulder Cabauw FINO-1 FINO-2 FINO-3 Hamburg Karlsruhe Los Alamos −2 2 4 6 8 10 12 14 time after sunset [h] 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

vSBL to wSBL transition

Transitions occur in about 40-70 % of nights

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 6 / 14

Frequency of transitions

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

wSBL to vSBL transition

Boulder Cabauw FINO-1 FINO-2 FINO-3 Hamburg Karlsruhe Los Alamos −2 2 4 6 8 10 12 14 time after sunset [h] 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

vSBL to wSBL transition

Transitions occur in about 40-70 % of nights wSBL to vSBL transitions are most probable around sunset whereas vSBL to wSBL transitions lack pronounced maxima

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 6 / 14

Frequency of transitions

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

wSBL to vSBL transition

Boulder Cabauw FINO-1 FINO-2 FINO-3 Hamburg Karlsruhe Los Alamos −2 2 4 6 8 10 12 14 time after sunset [h] 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 probability

vSBL to wSBL transition

Transitions occur in about 40-70 % of nights wSBL to vSBL transitions are most probable around sunset whereas vSBL to wSBL transitions lack pronounced maxima turbulence collapse and recovery events are in statistical equilibrium 4 hours after sunset

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 6 / 14

Frequency of timing between transitions

0.00 0.02 0.04 0.06 0.08 probability

Time between turbulence collapse to revovery

Boulder Cabauw FINO1 FINO2 FINO3 Hamburg Karlsruhe Los Alamos 2 4 6 8 10 12 time [h] 0.00 0.02 0.04 0.06 0.08 probability

Time between turbulence recovery to collapse

subsequent transitions occur in about 50-70 % of nights with transitions

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 7 / 14

Frequency of timing between transitions

0.00 0.02 0.04 0.06 0.08 probability

Time between turbulence collapse to revovery

Boulder Cabauw FINO1 FINO2 FINO3 Hamburg Karlsruhe Los Alamos 2 4 6 8 10 12 time [h] 0.00 0.02 0.04 0.06 0.08 probability

Time between turbulence recovery to collapse

subsequent transitions occur in about 50-70 % of nights with transitions the pdfs of the time between subsequent turbulence recovery or collapse events have a very similar shape independent of the tower station

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 7 / 14

Structure of the stratification during transitions

50 100 150 200 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 height [m]

vSBL to wSBL transitions

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 mean(Θh − Θ2) [K]

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 8 / 14

Structure of the stratification during transitions

50 100 150 200 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 height [m]

vSBL to wSBL transitions

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 mean(Θh − Θ2) [K]

radiative cooling leads to a steady formation of stable stratification suppressing vertical turbulent fluxes

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 8 / 14

Structure of the stratification during transitions

50 100 150 200 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 height [m]

vSBL to wSBL transitions

0.0 0.4 0.8 1.2 1.6 2.0 2.4 2.8 mean(Θh − Θ2) [K]

radiative cooling leads to a steady formation of stable stratification suppressing vertical turbulent fluxes very rapid break down of stratification allowing for enhanced turbulent vertical fluxes which stays quite steady after turbulence recovery

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 8 / 14

Structure of the wind shear during transitions

Along-wind component

50 100 150 200 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 height [m]

vSBL to wSBL transitions

0.00 1.25 2.50 3.75 5.00 6.25 7.50 8.75 10.00 mean(Uh parallel to U200) [m s−1]

Across-wind component

50 100 150 200 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 height [m]

vSBL to wSBL transitions

0.0 0.3 0.6 0.9 1.2 1.5 1.8 mean(Uh perpendicular to U200) [m s−1]

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 9 / 14

Structure of TKE during transitions

50 100 150 200 250 height [m]

wSBL to vSBL transitions

−80 −60 −40 −20 20 40 60 80 time [min] 50 100 150 200 250 height [m]

vSBL to wSBL transitions

0.200 0.275 0.350 0.425 0.500 0.575 0.650 0.725 mean(TKE) [m2 s−2]

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 10 / 14

Change of wind power density during times of transitions

−500 500 1000 1500 2000 ∂t0.5ρW3

80 [W m−2 s−1]

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 pdf wSBL → vSBL vSBL → wSBL

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 11 / 14

Tendencies of geostrophic forcing during transitions

−4 −2 2 4 ∂t Ugeo [m s−1 h−1] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 pdf wSBL to vSBL vSBL to wSBL

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 12 / 14

Tendencies of geostrophic forcing during transitions

−4 −2 2 4 ∂t Ugeo [m s−1 h−1] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 pdf wSBL to vSBL vSBL to wSBL

systematic changes of the geostrophic wind in times of transitions is absent

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 12 / 14

Evolution of low level cloud cover during transitions

20 40 60 80 100 low level cloud cover [%]

wSBL to vSBL transition

10-%ile 20-%ile 30-%ile 40-%ile 50-%ile 60-%ile 70-%ile 80-%ile −80 −60 −40 −20 20 40 60 80 time [min] 20 40 60 80 100 low level cloud cover [%]

vSBL to wSBL transition

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 13 / 14

Conclusions

The evolution of state variables in times of transition (as classified by an HMM analysis) is physically reasonable and shows clear structures during turbulence collapse and recovery events

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 14 / 14

Conclusions

The evolution of state variables in times of transition (as classified by an HMM analysis) is physically reasonable and shows clear structures during turbulence collapse and recovery events A clear precursor in external state variables is not always evident, however, changes in low level cloud cover have the potential to initiate a transition

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 14 / 14

Conclusions

The evolution of state variables in times of transition (as classified by an HMM analysis) is physically reasonable and shows clear structures during turbulence collapse and recovery events A clear precursor in external state variables is not always evident, however, changes in low level cloud cover have the potential to initiate a transition Substantial changes in the wind power density occur in times of transitions. With surface

• bservations HMM analyses can assess these effects for different sites around the world.
• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 14 / 14

Conclusions

The evolution of state variables in times of transition (as classified by an HMM analysis) is physically reasonable and shows clear structures during turbulence collapse and recovery events A clear precursor in external state variables is not always evident, however, changes in low level cloud cover have the potential to initiate a transition Substantial changes in the wind power density occur in times of transitions. With surface

• bservations HMM analyses can assess these effects for different sites around the world.

Reynolds-averaged state variables carry all important information about transitions allowing to work towards new stochastic boundary layer parameterisations of the two regimes in weather and climate models

• C. Abraham, A. H. Monahan (UVic)

June 14, 2018 14 / 14