Climate Extremes The Drought Hazard Bradfield Lyon International Research Institute for Climate and Society The Earth Institute, Columbia University US.

Climate ExtremesThe Drought Hazard

Bradfield LyonInternational Research Institute for Climate and Society

The Earth Institute, Columbia University

US CLIVAR Summit on Climate Extremes

Denver, CO 7-9 July 2010

US CLIVAR Drought Working Group

Drought in Coupled Models Project – “DRICOMP”

Drought Interest Group

• Difficult to define. Fundamentally, an insufficient supply of water to meet demand but demands are many, vary with region and sector, and supply can be of non-local origin (e.g., Tucson, AZ and the CO River)

• When does “drought” start? Terminate? As measured by what? Relevant to?

• Occurs on multiple timescales – often simultaneously (consider its impacts)

• In all cases, ultimately tied to “extended” periods of deficient precipitation relative to the “expected” value for a particular location but that includes:

- Late onset or early demise of monsoon rainfall- Monsoon breaks- Sub-seasonal SI multi-year multi-decadal CC

• Multiple causes. Linked to regional and large scale atmospheric circulation anomalies (some related to SSTs) and land surface-atmosphere interactions

• Enhanced drought prediction depends fundamentally on improved predictions of precipitation (and other variables related to surface fluxes of water and energy)

Drought – An Extreme Challenge

Figure: UN World Water Development Report-2, Chapter 4, Part 1. Global Hydrology and Water Resources

Monitoring Drought – What Aspect?

Monitoring Drought – Characteristic Time ScalesCorrelation Top Layer VIC Soil Moisture and SPI-3 (1950-2000)

All Months May-Sep only

VIC data Courtesy of

Justin Sheffield,Princeton Univ.

Hudson Valley NYStd. VIC SM1

Anomalies(monthly, 5-yr)

Monitoring DroughtNumerous “drought” indices in use each with its own “intrinsic” time scale

- PRCP -- monthly, past 90 days, water year, standardized indices (SPI)- Water balance indices: “P-E”, PDSI, etc.- Soil Moisture (typically modeled, experimental satellite products)- Snow Water Equivalent, Surface Water Supply Index- Streamflow - Vegetation Condition (satellite estimates) …

Challenges:

- Observational data are imperfect; scale issues (information, decisions)

- Lack of real time updates for monitoring and prediction (“preliminary”)

- Lack of long historical records for satellite-derived (and other) products

- Higher frequency (daily) precipitation also of interest but often unavailable

- Derived quantities (e.g. model soil moisture) subject to input uncertainties and observations for calibration/comparison are sparse

- Relevance of indices (and predictions) to specific applications -- the “best” drought index is the one most closely associated with the specific

application of interest (ag, rangeland condition, streamflow, etc.)

RMS Difference inMonthly PRCPGPCC – UEA

as a Fraction ofGPCC Annual Mean

(1971-2000)

CPCSPI-12 < -1.0

CPCSPI-12 > +1.0

e-folding time (months)

To = e-folding time for run durations in SPI-12

Slide Courtesy of Kingtse MoCPC Drought Briefing for May 2010

ModeledSoil Moisture Estimates(Runoff, ET, Soil Saturation)

• Derived variables influenced by uncertainties in model inputs and different model designs

• Use model-relative measures of variability (e.g., percentiles) for comparisons across models in near real time

• Need for enhanced observations of soil moisture

• Better flux measurements for comparison with models (Ameriflux)

Estimates of “Soil Moisture” from Satellite

SMOS – ESA; SMAP – NASA

SMOS

Figure: ESA

Prolonged Drought -- The Role of SSTs

Schubert et al., 2004

Seager et al., 2005

Prolonged Drought – The Role of SSTs

POGA-ML (similar to GOGA)

Prolonged Drought – “ENSO +”

Protracted Drought 1998-2002

Hoerling and Kumar, 2002

Trends: Coupled Models vs. AMIP

Shin and Sardeshmukh, 2010

1988 Drought

AMJ SST Anomaly

AMJ PRCP &250 hPa Std.

Hght. Anomalies

Reanalysis

CFSECHAM4.5

AMJ 1988Anomalous

Stationary Waves

SPI-6OBSJun 1988

SPI-6NSIPP Jun 1988

SPI-6ECHAM4.5 Jun 1988

Observations & AMIP Simulations: Drought of 1988

SPI-6GFDL 2.14 Jun 1988

GLACE-2 (GEWEX, CLIVAR)• Used best estimate of soil moisture from offline (similar to GSWP-2)• Compared control with initialized land sfc. runs across multiple GCMS (10)

Role of the Land Surface

Koster et al., 2010

Koster et al., 2010

Role of the Land Surface

Seneviratne et al., 2006

• Overall (global scale) soil moisture memory reasonably simulated• Regional biases important for the practical application of model output

Role of the Land Surface – Model Biases (GLACE)

Towards Probabilistic Prediction of Meteorological Drought

• Predictive information from both persistence and GCM

• AMIP -- Does not include the role of land surface condition

Towards Probabilistic Prediction of Meteorological Drought

Importance of the Sub-seasonal Time Scale:Dynamic Crop Models

• Account for dynamic, nonlinear crop-soil-weather interactions

• Need DAILY weather inputs in crop models

• Requires disaggregation of seasonal forecasts to obtain daily sequences of T, P

Ob

serv

ed y

ield

(kg

ha-1

)

Rainfall (mm day-1) Mean max temperature (°C) Simulated yield (kg ha-1)

Observed soybean yields (GA, USA yield trials) vs. seasonal rainfall, temperature, simulated yields

Slide Courtesy of James Hansen, IRI

0

10

20

30

40

50

60

70

80

90

1 16 31 46 61 76 91 106

121

136

151

166

181

196

211

226

241

256

271

286

301

316

331

346

361

0

10

20

30

40

50

60

70

80

90

1 16 31 46 61 76 91 106

121

136

151

166

181

196

211

226

241

256

271

286

301

316

331

346

361

Observed Rainfall

0

10

20

30

40

50

60

70

80

90

1 16 31 46 61 76 91 106

121

136

151

166

181

196

211

226

241

256

271

286

301

316

331

346

361

Bias-Corrected GCM:(Amplitude, Frequency)

Raw GCM Daily Rainfall:Amplitude, Frequency Bias

Ines and Hansen (2006), Hansen et al. (2006)

Bias-Corrected Daily Rainfall from a GCM

• GCM over-estimates the OBS autocorrelation of daily PRCP

• Changes in higher-frequency precipitation events of much interest to ag. and water sectors (including under CC)

Figure 11.12, IPCC AR4

Demand to Move Beyond One’s Means

Annual DJF JJA

0

2

4

6

8

10

12

14

165 6 6 7 7 7 8 8 8 9 9 9

10 10 11 11 12 13

> 30 mm

21-30 mm

11-20 mm

1-10 mm

“Near-Normal” Monthly Precipitation in Central Park(within +/- 5% of long term median value)

No. Days with Precipitation

Cum

. D

ays

in C

ateg

ory

DRAFT TOC, “DIG” Whitepaper on Drought

~ finis ~

FIG. 2. Scatterplot of the percentage change in global-mean column-integrated (a),(c) water vapor and (b),(d) precipitation vs the global-mean change in surface air temperature for the PCMDI AR4 models under the (a),(b) Special Report on Emissions Scenarios (SRES) A1B forcing scenario and (c),(d) 20C3M forcing scenario. The changes are computed as differences between the first 20 yr and last 20 yr of the twenty-first (SRES A1B) and twentieth (20C3M) centuries. Solid lines depict the rate of increase in column-integrated water vapor (7.5% K-1). The dashed line in (d) depicts the linear fit of P to T, which increases at a rate of 2.2% K-1.

From: UN World Water Development Report, 2003

Graphic: Third UN Water Development Report, World Water Assessment Report, 2009

http://upload.wikimedia.org/wikipedia/commons/f/f2/World_population_growth_%28lin-log_scale%29.png

Graphic from -- Groundwater: A global assessment of scale and significance, IWMI, 2007

FIG. 2. Scatterplot of the percentage change in global-mean column-integrated (a),(c) water vapor and (b),(d) precipitation vs the global-mean change in surface air temperature for the PCMDI AR4 models under the (a),(b) Special Report on Emissions Scenarios (SRES) A1B forcing scenario and (c),(d) 20C3M forcing scenario. The changes are computed as differences between the first 20 yr and last 20 yr of the twenty-first (SRES A1B) and twentieth (20C3M) centuries. Solid lines depict the rate of increase in column-integrated water vapor (7.5% K-1). The dashed line in (d) depicts the linear fit of P to T, which increases at a rate of 2.2% K-1.

-3

-2

-1

0

1

2

3

Sep

-51

Sep

-53

Sep

-55

Sep

-57

Sep

-59

Sep

-61

Sep

-63

Sep

-65

Sep

-67

Sep

-69

Sep

-71

Sep

-73

Sep

-75

Sep

-77

Sep

-79

Sep

-81

Sep

-83

Sep

-85

Sep

-87

Sep

-89

Sep

-91

Sep

-93

Sep

-95

Sep

-97

Sep

-99

1-S

ep

3-S

ep

5-S

ep

7-S

ep

0

5,000

10,000

15,000

20,000

25,000

30,000

35,000

SPI-12

Avg Q

Drought Working Group

US CLIVAR DWG -- Idealized SST Runs

• Work done in parallel with the Drought in Coupled Models Project (DRICOMP)