The Madden-Julian Oscillation and extreme precipitation in the contiguous United States Charles Jones Leila Carvalho 1, Jon Gottschalk 2, Wayne Higgins.

The Madden-Julian Oscillation and extreme

precipitation in the contiguous United States

Charles Jones

Leila Carvalho1,

Jon Gottschalk2, Wayne Higgins2

1University of California, Santa Barbara

2Climate Prediction Center (CPC/NCEP)

2

Outline

Motivation:

Extreme precipitation (CONUS, winter)

Madden-Julian Oscillation (MJO)

Forecast skill of extreme Precipitation

Relative value of deterministic forecasts

Spatiotemporal variations in extreme

Precipitation in the CONUS and the MJO

3

Extreme precipitation

USA

Extreme precipitation

Winter storm

Heavy snow

Snowpack

Spring/summer water

o Most important mode of tropical intraseasonal variations

o Time scales of 30 to 60 dayso Anomalies propagate

eastward along the tropical belt; phase speeds ~ 5 m s-1

o strong interaction with clouds, rain, surface winds and large-scale circulation

The Madden-Julian Oscillation

5

6

Anomalous upper level circulation (200-hPa)

Enhanced Convection in the western Pacific

Coupled Forced Rossby-Kelvin wave response

Rossby waves

-+ -

+

Midlatitude wave train

Kelvin wave -

+

Madden-Julian Oscillation (MJO) Life Cycle

7

7

Barlow et al. (2005) Jones et al. (2004)

Mo and Higgins (1998) Higgins et al (2000) Jones (2000) Bond and Vecchi (2003) Jones et al. (2004)

Carvalho et al. (2004) Liebmann et al. (2004) Jones et al. (2004)

Wheeler and Hendon (2004) Jones et al. (2004)

Jones et al. (2004)

Jones, C., D. E. Waliser, K. M. Lau, and W. Stern, 2004: Global occurrences of extreme precipitation events and the Madden-Julian Oscillation: observations and predictability. J. Climate, 17, 4575-4589.

NH winter

The MJO and Extreme Precipitation: Observations

8

Madden-Julian Oscillation (MJO): influence on forecast skill and subseasonal predictability

MJO

o Waliser, D. E., K. M. Lau, W. Stern, and C. Jones, 2003: Potential predictability of the Madden-Julian Oscillation. Bull. American Meteo. Soc., 33-50.

o Jones, C., D. E. Waliser, K. M. Lau, and W. Stern, 2004: Global occurrences of extreme precipitation events and the Madden-Julian Oscillation: observations and predictability. J. Climate, 17, 4575-4589.

o Jones, C., D. E. Waliser, K. M. Lau, and W. Stern, 2004: The Madden-Julian Oscillation and its Impact on Northern Hemisphere Weather Predictability. Mon. Wea. Rev., 132, 1462-1471.

9

Data and Methodology

Definition of extreme precipitation Observations: daily gridded precipitation over the

contiguous United States, 1 Nov-31 Mar, 1981-2008. Model:

NCEP CFS model: CMIP run (30 yrs) NCEP CFS reforecasts (version 1)

Two levels of extreme precipitation: daily total exceeds the 75th and 90th percentiles of monthly pdf.

Madden-Julian Oscillation (MJO): influence on forecast skill of extreme precipitation

Jones, C., J. Gottschalck, L. M. V. Carvalho, and W. Higgins, 2010: Influence of the Madden-Julian Oscillation on forecasts of extreme precipitation in the contiguous United States. Monthly Weather Review (In press).

10

NCEP CFS Model Reforecasts (v1)

CFS Reforecasts

IC

9-mo forecasts

IC

o 15 initial conditions per montho Forecasts out to 270 days (CFS seasonal forecast

calibration)o Analyzed: forecasts out to 4 weeks

For each forecast:o Remove mean model bias in precipitation fieldo Forecasts of extremes when precipitation exceeds

adjusted 75th or 90th percentiles o Adjustment: ratios between observations and CFS CMIP

run percentilesAnalyzed deterministic and probabilistic forecast skill

11

NCEP/NCAR reanalysis: U200, U850 intraseasonal anomalies

combined EOF

Phase diagram from PC1/PC2

MJO event has amplitude > 0.9

Phase rotates anti-clockwise

70 MJO events during 1 Nov-31 Mar, 1981-2008

(phases ~Wheeler and Hendon 2004)

Enhanced convection

Identification of MJO

12

Non-Probabilistic Forecasts: yes/no

YES NO

YES a b

NO c d

Observations

Fo

rec

as

t

2x2 Contingency Table

n= a+b+c+d forecasts

a

c

b

d

13

oTwo samples: active MJO versus inactive (null)oVerification on MJO phases (≠ than MJO state in ICs)oStatistical significance: comparison of scores in active

MJO with scores from resampled inactive cases

Deterministic forecasts

14

lead times (days)

Forecast skill during active MJO90th percentile precipitation extreme

Mean Heidke skill scores (HSS) over gridpoints that are significant at 5% level relative to inactive MJO conditionsMin

Max

lead times (days)

Min

Max

Relative value of deterministic forecasts

YES NO

YES a b

NO c d

Event occurs

Ac

tio

n t

ak

en

Decision makerL= loss if event occurs

and no action takenC= cost if action taken

L 0

C C

Jones, C., L. M. V. Carvalho, J. Gottschalck and W. Higgins, 2010: The Madden-Julian Oscillation and the relative value of deterministic forecasts of extreme precipitation in the contiguous United States. J. Climate (in press) 

16

Cost/loss ratio decision model

Where V is value and: o = user’s cost/loss ratio (C/L)o s = climatological base rate of the event (90th extreme)o H = hit rateo F = false alarm rate

When = s potential forecast value

Caveats: simple model, user’s actions expressed numerically, assume users are risk neutral-only concerned with long-term average expense

Evaluated here with deterministic forecasts

17

MJO

Summary: maximum value over USA: MJO x null

18

Spatiotemporal variations in extreme precipitation in the contiguous United States and

the Madden-Julian OscillationCharles Jones and Leila M. V. Carvalho

Objective: develop a probabilistic approach to spatiotemporal variations in extreme precipitation in the CONUS during and relationships with the MJO

Data and Methodology Data: NCEP CPC Unified gridded precipitation,

0.5 lat/lon, 1 Nov – 31 Mar 1979-2010 Definition of extreme precipitation: daily average

exceeds 90th percentile of monthly pdf.

19

Example of extreme precipitation

20

21

Contiguous Region of Extreme Precipitation (CREP)

CREP: regions of spatially connected gridpoints in which daily precipitation exceed the 90th percentile

Properties: mean precipitation, area, orientation, elongation, frequency of occurrence during active MJO and inactive days

Probabilities of CREPs occurrence given status of MJO, anomalies in geopotential height at 500-hPa

22

Counts of CREPS during active MJO (top) and quiescent days (bottom). Counts were assigned to centers of each CREP during 1 November-31 March, 1979-2010. Total number of CREPs: 5600. Percentages of CREPS during active and inactive MJO days are indicated in parenthesis in each panel. Maximum number of counts are indicated on the left corner of each panel.

23

Spatial and size distributions of CREPs during active MJO days. Circles indicate locations of centers of CREPs in each MJO phase; phases are shown on the bottom right corner of each panel. Sizes of circles represent areas (km2); scale shown at the bottom. Total number of CREPs: 3869. Percentages of CREPs in each MJO phase are shown on the bottom left of each panel.

24

January 1995 case study

Dates

MJO phases

Centers of negative H500 anomalies

CREPs

Distances from center of H500 in which there is a probability Pb of occurring one or more CREPs

cost/loss ratio (C/L)

Value of forecasts of extreme precipitation

cost/loss ratio (C/L)cost/loss ratio (C/L)

Curves show value of forecasts averaged on 1-3, 5-7, 9-11 and 12-14 days lead times.

Val

ue

Val

ue

Val

ue

MJO Phase 8

MJO Phase 1

Null

2626

26

Schematic representation of eastward propagation of tropical convective anomalies associated with the Madden-Julian Oscillation, changes in the Jet Stream and precipitation events in the Pacific Northwest.

Tropical-Extratropical Interaction associated with Madden-Julian

Oscillation