Convection-resolving climate change simulations: Short-term - - PowerPoint PPT Presentation

▶
convection resolving climate change simulations short
SMART_READER_LITE
LIVE PREVIEW

Convection-resolving climate change simulations: Short-term - - PowerPoint PPT Presentation

Oct 24, 2023 111 likes •564 views

Convection-resolving climate change simulations: Short-term precipitation extremes in a changing climate Nikolina Ban 1 , urg Schmidli 2 and Christoph Sch ar 1 J 1 Institute for Atmospheric and Climate Science, ETH Z urich 2 Goethe

slide-1
SLIDE 1

Convection-resolving climate change simulations: Short-term precipitation extremes in a changing climate

Nikolina Ban1, J¨ urg Schmidli2 and Christoph Sch¨ ar1

1Institute for Atmospheric and Climate Science, ETH Z¨

urich

2Goethe University, Frankfurt

8th RegCM Workshop May 2016, ICTP, Trieste

slide-2
SLIDE 2

Introduction Method Evaluation Climate Change crCLIM Summary

Hydrological Impacts of Heavy Precipitation Flash floods

Saanen (Switzerland), Jul 2010

Landslides

Graub¨ unden (Switzerland), Aug 2014

2 Nikolina Ban : CRM climate simulations

slide-3
SLIDE 3

Introduction Method Evaluation Climate Change crCLIM Summary

Link Between Temperature Change and Extreme Precipitation Change

Percentile Intensity [mm/d] Increase [%]

+7%/K

Daily precipitation

[Allen and Ingram, 2002]

3 Nikolina Ban : CRM climate simulations

slide-4
SLIDE 4

Introduction Method Evaluation Climate Change crCLIM Summary

Link Between Temperature Change and Extreme Precipitation Change

Percentile Intensity [mm/d] Increase [%]

+7%/K

Daily precipitation

[Allen and Ingram, 2002] Temperature Intensity [mm/h]

Hourly precipitation

[Lenderink and van Meijgaard, 2008]

  • Do heavy hourly precipitation events increase at adiabatic (∼6-7 %/K) or

super-adiabatic (∼14 %/K) rate?

3 Nikolina Ban : CRM climate simulations

slide-5
SLIDE 5

Introduction Method Evaluation Climate Change crCLIM Summary

Numerical modeling of climate

Global climate model 100 km Regional climate model 25 km Convection-resolving model 1 km

  • CRM: Convection-resolving model enables explicit simulation of convection

(e.g., thunderstorms, rain showers) [Figures: E. Zubler]

4 Nikolina Ban : CRM climate simulations

slide-6
SLIDE 6

Introduction Method Evaluation Climate Change crCLIM Summary

Numerical modeling of climate

Global climate model 100 km Regional climate model 25 km Convection-resolving model 1 km

  • CRM: Convection-resolving model enables explicit simulation of convection

(e.g., thunderstorms, rain showers)

  • CRM pioneering studies: Grell et al., 2000; Hohenegger et al., 2008; Knote et

al., 2010; Kendon et al., 2012, 2014; Langhans et al., 2013; Prein et al., 2013; Rasmussen et al., 2014; Ban et al., 2014, 2015; Prein et al., 2015 (review paper), Brisson et al., 2016 [Figures: E. Zubler]

4 Nikolina Ban : CRM climate simulations

slide-7
SLIDE 7

Introduction Method Evaluation Climate Change crCLIM Summary

Objectives

Evaluation

  • Does CRM improve representation of precipitation distribution and

statistics?

  • How do precipitation extremes scale with temperature? With

Clausius-Clapeyron relation? Climate Change

  • Difference between CRM and conventional climate models?
  • Link between temperature change & precipitation change?

Continental-scale convection-resolving climate simulations (crCLIM)

5 Nikolina Ban : CRM climate simulations

slide-8
SLIDE 8

Introduction Method Evaluation Climate Change crCLIM Summary

Setup

Two-step one-way nesting: BC ⇒ CPM12 ⇒ CRM2

  • CPM12 and CRM2 use COSMO-CLM

v4.14

  • Boundary Conditions: ERA-Interim

reanalysis & MPI-ESM-LR (RCP8.5)

  • CPM12: Convection–Parameterizing

Model

  • △x=12 km (0.11◦)
  • XxYxZ=260x228x60
  • CRM2: Convection–Resolving Model
  • △x=2.2 km (0.02◦)
  • XxYxZ=500x500x60

6 Nikolina Ban : CRM climate simulations

slide-9
SLIDE 9

Introduction Method Evaluation Climate Change crCLIM Summary

Setup

Two-step one-way nesting: BC ⇒ CPM12 ⇒ CRM2

  • CPM12 and CRM2 use COSMO-CLM

v4.14

  • Boundary Conditions: ERA-Interim

reanalysis & MPI-ESM-LR (RCP8.5)

  • CPM12: Convection–Parameterizing

Model

  • △x=12 km (0.11◦)
  • XxYxZ=260x228x60
  • Parametrization of convection:

Tiedtke

  • CRM2: Convection–Resolving Model
  • △x=2.2 km (0.02◦)
  • XxYxZ=500x500x60
  • Convection explicitly resolved
  • Shallow convection: Tiedtke

6 Nikolina Ban : CRM climate simulations

slide-10
SLIDE 10

Introduction Method Evaluation Climate Change crCLIM Summary

Setup

Two-step one-way nesting: BC ⇒ CPM12 ⇒ CRM2

  • CPM12 and CRM2 use COSMO-CLM

v4.14

  • Boundary Conditions: ERA-Interim

reanalysis & MPI-ESM-LR (RCP8.5)

  • CPM12: Convection–Parameterizing

Model

  • △x=12 km (0.11◦)
  • XxYxZ=260x228x60
  • Parametrization of convection:

Tiedtke

  • CRM2: Convection–Resolving Model
  • △x=2.2 km (0.02◦)
  • XxYxZ=500x500x60
  • Convection explicitly resolved
  • Shallow convection: Tiedtke

The numerical simulations have been performed on the CRAY XT5 and CRAY XE6 at CSCS 6 Nikolina Ban : CRM climate simulations

slide-11
SLIDE 11

Introduction Method Evaluation Climate Change crCLIM Summary

Experiments: CRM Simulations for the Greater Alpine Region

2050 2100 2000 1950 Mean over 2081-2100 Temperature Change Year [IPCC AR5]

7 Nikolina Ban : CRM climate simulations

slide-12
SLIDE 12

Introduction Method Evaluation Climate Change crCLIM Summary

Experiments: CRM Simulations for the Greater Alpine Region

2050 2100 2000 1950 Mean over 2081-2100

Hindcast

Temperature Change Year ERA-Interim [IPCC AR5]

7 Nikolina Ban : CRM climate simulations

slide-13
SLIDE 13

Introduction Method Evaluation Climate Change crCLIM Summary

Experiments: CRM Simulations for the Greater Alpine Region

2050 2100 2000 1950 Mean over 2081-2100

CTRL Hindcast

Temperature Change Year ERA-Interim GCM [IPCC AR5]

7 Nikolina Ban : CRM climate simulations

slide-14
SLIDE 14

Introduction Method Evaluation Climate Change crCLIM Summary

Experiments: CRM Simulations for the Greater Alpine Region

2050 2100 2000 1950 Mean over 2081-2100

CTRL SCEN Hindcast

Temperature Change Year ERA-Interim GCM GCM [IPCC AR5]

7 Nikolina Ban : CRM climate simulations

slide-15
SLIDE 15

Introduction Method Evaluation Climate Change crCLIM Summary

Experiments: CRM Simulations for the Greater Alpine Region

2050 2100 2000 1950 Mean over 2081-2100

CTRL SCEN Hindcast

Temperature Change Year ERA-Interim GCM GCM

  • Wallclock time: 1×10y CRM2 → ≈4-8months

[IPCC AR5]

7 Nikolina Ban : CRM climate simulations

slide-16
SLIDE 16

Introduction Method Evaluation Climate Change crCLIM Summary

Evaluation of Precipitation in Present-Day Climate

  • ERA-Interim driven simulations (1998–2007)

8 Nikolina Ban : CRM climate simulations

slide-17
SLIDE 17

Introduction Method Evaluation Climate Change crCLIM Summary

The 90th percentiles of daily/hourly precipitation in JJA

The 90th percentiles of daily precipitation

[Obs - APGD (Isotta et al., 2014), EOBS (Haylock et al., 2008) and RdisaggH (W¨ uest et al., 2010)] (Ban et al., 2014 JGR)

9 Nikolina Ban : CRM climate simulations

slide-18
SLIDE 18

Introduction Method Evaluation Climate Change crCLIM Summary

The 90th percentiles of daily/hourly precipitation in JJA

The 90th percentiles of daily precipitation The 90th percentiles of hourly precipitation

[Obs - APGD (Isotta et al., 2014), EOBS (Haylock et al., 2008) and RdisaggH (W¨ uest et al., 2010)] (Ban et al., 2014 JGR)

9 Nikolina Ban : CRM climate simulations

slide-19
SLIDE 19

Introduction Method Evaluation Climate Change crCLIM Summary

Frequency Distribution of Precipitation (JJA)

1 5 10 15 20 25 30 35 40 0.05 0.1 0.2 0.3 0.5 1.0 precipitation [mm/day] Normalized frequency

  • bs

CRM2 CPM12

a) JJA | pD

[Analysis for 62 Swiss stations] (Ban et al., 2015 GRL)

10 Nikolina Ban : CRM climate simulations

slide-20
SLIDE 20

Introduction Method Evaluation Climate Change crCLIM Summary

Frequency Distribution of Precipitation (JJA)

1 5 10 15 20 25 30 35 40 0.05 0.1 0.2 0.3 0.5 1.0 precipitation [mm/day] Normalized frequency

  • bs

CRM2 CPM12

a) JJA | pD

1 5 10 15 20 0.05 0.1 0.2 0.3 0.5 1.0 precipitation [mm/h]

  • bs

CRM2 CPM12

b) JJA | pHmax

[Analysis for 62 Swiss stations] (Ban et al., 2015 GRL)

10 Nikolina Ban : CRM climate simulations

slide-21
SLIDE 21

Introduction Method Evaluation Climate Change crCLIM Summary

Evolution of the Hourly Precipitation (July 12-14, 2006)

Obs → combined radar and rain gauge observations (W¨ uest et al., 2010) CRM2 → explicit convection (△=2.2km) CPM12 → parametrized convection (△=12km)

(Ban et al., 2014 JGR)

11 Nikolina Ban : CRM climate simulations

slide-22
SLIDE 22

Introduction Method Evaluation Climate Change crCLIM Summary

Diurnal Cycle of Summer Precipitation

Mean precipitation Wet-hour frequency Heavy precipitation [Analysis for 62 Swiss stations]

  • CRM2 realistically simulates amplitude and phase of the diurnal cycle

(Ban et al., 2015 GRL)

12 Nikolina Ban : CRM climate simulations

slide-23
SLIDE 23

Introduction Method Evaluation Climate Change crCLIM Summary

Scaling of Extreme Hourly Precipitation Events

· · · 7% increase per ◦C (as Clausius-Clapeyron)

  • - - 14% increase per ◦C

(Ban et al., 2014 JGR)

13 Nikolina Ban : CRM climate simulations

slide-24
SLIDE 24

Introduction Method Evaluation Climate Change crCLIM Summary

Scaling of Extreme Hourly Precipitation Events

· · · 7% increase per ◦C (as Clausius-Clapeyron)

  • - - 14% increase per ◦C
  • Super-adiabatic scaling captured by both models

(Ban et al., 2014 JGR)

14 Nikolina Ban : CRM climate simulations

slide-25
SLIDE 25

Introduction Method Evaluation Climate Change crCLIM Summary

Projections of precipitation

  • based on GCM-driven scenarios for 2081-2090 (RCP8.5) versus 1991-2000

15 Nikolina Ban : CRM climate simulations

slide-26
SLIDE 26

Introduction Method Evaluation Climate Change crCLIM Summary

Summer precipitation

Relative change → SCEN−CTRL

CTRL

∆Mean CRM2

(Ban et al., 2015 GRL)

16 Nikolina Ban : CRM climate simulations

slide-27
SLIDE 27

Introduction Method Evaluation Climate Change crCLIM Summary

Summer precipitation

Relative change → SCEN−CTRL

CTRL

∆Mean CRM2 ∆p99.99H

↑

2 events per 10 seasons

  • Increase in extreme precipitation despite an overall drying

(Ban et al., 2015 GRL)

16 Nikolina Ban : CRM climate simulations

slide-28
SLIDE 28

Introduction Method Evaluation Climate Change crCLIM Summary

Summer Precipitation on Daily Timescales

Relative change in percentile intensities → (SCEN-CTRL)/CTRL

[Average across the CRM2 domain]

  • Close agreement of CRM2 and CPM12

(Ban et al., 2015 GRL)

17 Nikolina Ban : CRM climate simulations

slide-29
SLIDE 29

Introduction Method Evaluation Climate Change crCLIM Summary

Summer Precipitation on Hourly Timescales

[Average across the CRM2 domain] (Ban et al., 2015 GRL)

18 Nikolina Ban : CRM climate simulations

slide-30
SLIDE 30

Introduction Method Evaluation Climate Change crCLIM Summary

Summer Precipitation on Hourly Timescales

[Average across the CRM2 domain]

  • CRM2 exhibits smaller changes than CPM12

(Ban et al., 2015 GRL)

18 Nikolina Ban : CRM climate simulations

slide-31
SLIDE 31

Introduction Method Evaluation Climate Change crCLIM Summary

Link Between Temperature Change and Extreme Precipitation Change

Moistening of the atmosphere is determined by Clausius-Clapeyron relation: 1 esat desat dT ≈ 6 − 7%/K = ⇒ 1 Pextreme dPextreme dT ≈ 6 − 7%/K

19 Nikolina Ban : CRM climate simulations

slide-32
SLIDE 32

Introduction Method Evaluation Climate Change crCLIM Summary

Link Between Temperature Change and Extreme Precipitation Change

Moistening of the atmosphere is determined by Clausius-Clapeyron relation: 1 esat desat dT ≈ 6 − 7%/K = ⇒ 1 Pextreme dPextreme dT ≈ 6 − 7%/K Daily precipitation (JJA) 6.5%/K

(Ban et al., 2015)

19 Nikolina Ban : CRM climate simulations

slide-33
SLIDE 33

Introduction Method Evaluation Climate Change crCLIM Summary

Link Between Temperature Change and Extreme Precipitation Change

Moistening of the atmosphere is determined by Clausius-Clapeyron relation: 1 esat desat dT ≈ 6 − 7%/K = ⇒ 1 Pextreme dPextreme dT ≈ 6 − 7%/K Daily precipitation (JJA) 6.5%/K Hourly precipitation (JJA) 6.5%/K ⇒ Extreme daily and hourly precipitation asymptotically intensify with the Clausius-Clapeyron relation

  • Assessment uses all-event percentiles (Sch¨

ar et al., 2016)(Ban et al., 2015)

19 Nikolina Ban : CRM climate simulations

slide-34
SLIDE 34

Introduction Method Evaluation Climate Change crCLIM Summary

Convection-Resolving Climate Modeling on Future Supercomputing Platforms (crClim)

http://www.c2sm.ethz.ch/research/crCLIM.html

20 Nikolina Ban : CRM climate simulations

slide-35
SLIDE 35

Introduction Method Evaluation Climate Change crCLIM Summary

European-Scale Convection-Resolving Climate Simulations (crClim)

  • Two-step one-way nesting:

ERA-Interim ⇒ 12km ⇒ 2.2km

  • 1536×1536×60 grid points
  • 10-year long period: 1999-2008

⇒ Completed

  • Wall-clock time: 1 year ⇒ 5 days
  • GPU version of COSMO (Fuhrer et

al., 2014)

.

CTRL2 dx = 0.02° ≈ 2.2km CTRL12 dx = 0.11° ≈ 12km Ban et al. 2014 K e n d

  • n

e t a l . 2 1 2

(Leutwyler et al., 2016 Submitted to GMD)

21 Nikolina Ban : CRM climate simulations

slide-36
SLIDE 36

Introduction Method Evaluation Climate Change crCLIM Summary

European-Scale Convection-Resolving Climate Simulations (crClim)

  • Two-step one-way nesting:

ERA-Interim ⇒ 12km ⇒ 2.2km

  • 1536×1536×60 grid points
  • 10-year long period: 1999-2008

⇒ Completed

  • Wall-clock time: 1 year ⇒ 5 days
  • GPU version of COSMO (Fuhrer et

al., 2014)

  • Dynamical core rewritten in C++
  • Parameterizations use OpenACC
  • Runs on Piz Daint (Cray XC30,

CSCS)

  • Used for operational NWP at

MeteoSwiss (∆x=1 km)

.

CTRL2 dx = 0.02° ≈ 2.2km CTRL12 dx = 0.11° ≈ 12km Ban et al. 2014 K e n d

  • n

e t a l . 2 1 2

(Leutwyler et al., 2016 Submitted to GMD)

21 Nikolina Ban : CRM climate simulations

slide-37
SLIDE 37

Introduction Method Evaluation Climate Change crCLIM Summary

Diurnal Cycle of Convection

mm/h

(David Leutwyler)

22 Nikolina Ban : CRM climate simulations

slide-38
SLIDE 38

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation

23 Nikolina Ban : CRM climate simulations

slide-39
SLIDE 39

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation
  • CRM2 exhibits super-adiabatic and adiabatic scaling for hourly

warm-season precipitation, while only adiabatic for hourly cold-season precipitation (in accordance with observations)

23 Nikolina Ban : CRM climate simulations

slide-40
SLIDE 40

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation
  • CRM2 exhibits super-adiabatic and adiabatic scaling for hourly

warm-season precipitation, while only adiabatic for hourly cold-season precipitation (in accordance with observations)

  • Close agreement of CRM2 and CPM12 regarding the changes in daily

precipitation; for hourly extreme precipitation CRM2 exhibits smaller changes than CPM12

23 Nikolina Ban : CRM climate simulations

slide-41
SLIDE 41

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation
  • CRM2 exhibits super-adiabatic and adiabatic scaling for hourly

warm-season precipitation, while only adiabatic for hourly cold-season precipitation (in accordance with observations)

  • Close agreement of CRM2 and CPM12 regarding the changes in daily

precipitation; for hourly extreme precipitation CRM2 exhibits smaller changes than CPM12

  • CRM2 is consistent with theoretical expectations ⇒ Changes in extreme

summer precipitation qualitatively scale with the Clausius-Clapeyron rate

23 Nikolina Ban : CRM climate simulations

slide-42
SLIDE 42

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation
  • CRM2 exhibits super-adiabatic and adiabatic scaling for hourly

warm-season precipitation, while only adiabatic for hourly cold-season precipitation (in accordance with observations)

  • Close agreement of CRM2 and CPM12 regarding the changes in daily

precipitation; for hourly extreme precipitation CRM2 exhibits smaller changes than CPM12

  • CRM2 is consistent with theoretical expectations ⇒ Changes in extreme

summer precipitation qualitatively scale with the Clausius-Clapeyron rate ⋆ Currently this work is extended to simulations that cover Europe

23 Nikolina Ban : CRM climate simulations

slide-43
SLIDE 43

Introduction Method Evaluation Climate Change crCLIM Summary

Summary

  • CRM2 strongly improves the simulation of the sub-daily precipitation
  • CRM2 exhibits super-adiabatic and adiabatic scaling for hourly

warm-season precipitation, while only adiabatic for hourly cold-season precipitation (in accordance with observations)

  • Close agreement of CRM2 and CPM12 regarding the changes in daily

precipitation; for hourly extreme precipitation CRM2 exhibits smaller changes than CPM12

  • CRM2 is consistent with theoretical expectations ⇒ Changes in extreme

summer precipitation qualitatively scale with the Clausius-Clapeyron rate ⋆ Currently this work is extended to simulations that cover Europe Thank you for your attention!

23 Nikolina Ban : CRM climate simulations