The ET-CCDI Indices as a
starting point for ET-SCI
Enric Aguilar,
Center for Climate Change, C3, URV, Tarragona, Spain.
THE PROBLEM
(BACK TO THE 1990s)
• The Second Assessment Report (SAR) of the
Intergovernmental Panel on Climate Change (IPCC),
published in 1996, concluded that the available data
and analyses were inadequate for any assessment to
be made about the nature of global changes in
extreme climate events.
• Because of the traditional focus of climatologists on
monthly data and the proprietary view many
countries have toward data on shorter time scales, the
international exchange of long-term daily climate
records has been limited, and in 1996 there was no
international dataset of long-term daily terrestrial
data available
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
2
THE ETCCDI SOLUTION (I)
• In 1999 what is now known as the joint Expert Team on
Climate Change Detection and Indices (ETCCDI) was
organized
• The expert team is jointly sponsored by the World
Meteorological Organization (WMO) Commission for
Climatology (CCl), the World Climate Research
Programme (WCRP) project on Climate Variability and
Predictability (CLIVAR) and, since 2006, the Joint WMOIntergovernmental Oceanographic Commission (IOC) of
the United Nations Educational, Scientific and Cultural
Organization (UNESCO) Technical Commission for
Oceanography and Marine Meteorology (JCOMM).
www.clivar.org/organization/etccdi/etccdi.php
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
3
Commision for
Climatology and
ETCCDI
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
4
THE ETCCDI SOLUTION (II)
• ETCCDI Goals:
• Internationally coordinating the exact formulation of a
suite of agreed indices of climate extremes from daily
data + software to compute, allowing international
comparison
• Promote the analysis of extremes around the world by
organizing regional climate-change workshops following
the model pioneered in December 1998 at an AsiaPacific Network for Global Change Research (APN)
meeting in Melbourne: local experts + international
experts = capacity development + scientific papers.
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
5
WORKSHOPS
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
6
... IN THE CARIBBEAN …
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
7
IMPACT OF ETCCDI WORKSHOPS (and
similar initiatives) OVER GLOBAL
CLIMATE EXTREMES KNOWLEDGE
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
8
WHY THE ETCCDI INDICES
• Indices are not data, easier exchange
• Based on daily temperature and daily precipitation,
the most commonly available climate data
• Light extremes: they look at events that happen a
few times a year, making “easier” trend analysis
(i.e. Looking at the 90th percentile instead to at the
99th percentile)
• Their formulation allows to understand how
climate changes not only how the mean climate
changes.
2/16
WMO Workshop on Enhancing Climate Indices
for Sector-Specific Applications in the
Caribbean region
9
Approach to calculate climate indices
Based on fixed threshold value:
•
•
summer days (days with tmax > 25°C)
heavy precipitation days (days with prec > 10 mm)
• Easy to understand but not significant to all regions of the world:
•
ice days (days with tmax < 0°C) are not too frequent in the Caribbean!
Based on variable threshold value:
•
•
warm nights (% days when tmin > 90th percentile)
very wet days (total prec when prec > 95th percentile)
•
More difficult to interpret but facilitate comparison between different parts
of the world
Calculation of the percentiles
Minimum temperature at De Bilt
20
20
Temperature indices
•
•
Temperature (°C)
90th percentile
10th & 90th percentiles calculated from 1961-1990
percentiles obtained for each day of the year using a 5-day
window centered at the calendar day Jones et al. 1999
new approach based on bootstrap methodology
developed by Zhang et al. 2004 to obtain percentiles
within the base period to provide temporally consistent
estimate of threshold in and out the base period
10
10
0
0
10th percentile
-10
-10
-20
-20
1
2
3
4
5
6
7
8
9
11
12
Jones et al. 1999
Month
Days above the 99th percentile
Average of all stations in Canada
10
with bootstrap
Precipitation indices
•
95th & 99th percentiles calculated from all days
during the year when prec >1 mm during 19611990
without bootstrap
Zhang et al. 2004
Missing values
•
indices are calculated on
•
•
•
indices calculated on monthly basis
•
•
monthly and annual bases
annual basis only
if number of days missing > 3 days then monthly value missing
indices calculated on annual basis
•
if number of days missing > 15 days or monthly value is missing
then annual value missing
Definition of temperature indices
(16 indices)
Cold Extremes
Warm Extremes
Frost days (tmin < 0°C)
Ice days (tmax < 0°C)
Monthly lowest value in tmax
Monthly lowest value in tmin
Cold nights (% days w tmin < 10th perc.)
Cold days (% days w tmax < 10th perc.)
Cold spell duration index (count of days w
at least 6 cons. days w tmin > 10th perc.)
Summer days (tmax > 25°C)
Tropical nights (tmin > 20°C)
Monthly highest value in tmax
Monthly highest value in tmin
Warm nights (% days w tmin > 90th perc.)
Warm days (% days w tmax > 90th perc.)
Warm spell duration index (count of days w
at least 6 cons. days w tmax > 90th perc.)
Others
Growing season length (6 days with TG >5°C & 6 days
with TG < 5°C; North & South Hemispheres)
Diurnal temperature range (monthly mean difference
between tmax & tmin
Example
Definition of precipitation indices
(11 indices)
Extremes
Monthly highest 1-day prec
Monthly highest 5-day cons. prec
Heavy prec days (prec > 10 mm)
Very heavy prec days (prec > 20 mm)
Consecutive dry days (max number of
cons. days w prec < 1 mm)
Consecutive wet days (max number of
cons. days w prec ≥ 1 mm)
Very wet days (annual total prec w
prec > 95th perc.)
Extremely wet days (annual total prec w
prec > 99th perc.)
Others
Simple Day Intensity Index (total prec
divided by number of wet days)
Days w prec > xx mm
Annual total precipitation
Example
Thanks!