Climate variability and climate model application in Indonesia

1
Climate variability and climate model application
in Indonesia

• Edvin Aldrian
• Agency for the Assessment and Application of
  Technology (BPPT), Indonesia
• Marine Study Program, University of Indonesia,
  Jakarta
• Email edvin_at_webmail.bppt.go.id
• http//geocities.com/e_aldrian

The Institute of Asian Monsoon System Prediction
of Change and Variability Honolulu, 3 January 2008
2
Climatology of regional rainfall

• Mainly monsoonal
• Three distinct rainfall climate regions

Aldrian and Susanto, 2003, Intl J Climatol.
monsoonal
Semi-monsoonal
Anti-monsoonal
3
Interannual variability in Region A
Aldrian et al, 2007, Theo. Appl. Climatol.
4
Interannual variability in Region B
Aldrian et al, 2007, Theo. Appl. Climatol.
5
Interannual variability in Region C
Aldrian et al, 2007, Theo. Appl. Climatol.
6
Wide range SST responses to rainfall in region A

• Weak response in spring, no ENSO influence
• Strong two dipoles in SON (Walker cell)
• Role of SPCZ in SON

Aldrian and Susanto, 2003, Intl J Climatol.
7
Wide range SST responses to rainfall in region B

• Weak response in all season, especially in
  spring
• no ENSO influence and walker cell

Aldrian and Susanto, 2003, Intl J Climatol.
8
Wide range SST responses to rainfall in region C

• Weak response in spring
• Strong two dipoles in SON (Walker cell) like
  region A
• Role of SPCZ in SON like in region A

Aldrian and Susanto, 2003, Intl J Climatol.
9
Skills of monthly rainfall variability
Response to ENSO

• Negative responses to NINO3 SST
• Significant in MJJAS
• (similar to Hendon, 2003)

Variability in comparison to observations (correla
tion values)

• Strong responses in Region A and C
• Spring is the least responsive season
• ECHAM4 responds well to ENSO

Aldrian et al, 2007, Theo. Appl. Climatol.
10
Several finished and on going applications of
climate modeling

• Rainfall simulation (dissertation project)
• Forest Fire smoke distribution (INSIDE project).
• An EU Asia ProEco Project with a collaboration
  between BPPT, MPI and Univ. Cambridge
• Local ocean model simulation (LITHMOS project)
• A proposed DFG Project with a collaboration
  between BPPT, IfM Uni Hamburg and Uni Syiah Kuala
  of Aceh Indonesia

11
The atmosphere and ocean model

• REMO ECHAM Physics
• allows only one surface type in each grid cell
• 0.50 101x55 grid cells for entire Indonesia,
• 1/60 73x61 grid cell over Sulawesi
• 20 layers
• boundary forcing every 6 hr
• ERA15 1979-1993,
• NCEP/NCAR 1979-1993,
• ECHAM4 1979-1988
• Higher resolution product
• allow a detailed investigation
• of five large islands
• and three sea regions

• MPI OM1
• Using a conformal grid system with desired pole
  locations
• North pole 112E 29N
• South pole 132E 22S
• Two model resolutions 182x105 and 362x210
• 20 layers for low resolution, 30 layers for high
  resolution
• highest resolution over the Maritime Continent
• boundary forcings every 6 hr
• ERA15 1979-1993,
• NCEP/NCAR 1948-1998,
• OMIP climatology

12
Improvement by higher resolution with Regional
Model
Aldrian et al, 2004, Clim Dyn.
These two examples are taken from a normal or
non-ENSO year
13
Rainfall Predictability by Regional Model
Aldrian et al, 2004, Clim Dyn.
RMS Error against control run
Ensemble runs with REMO-ERA and REMO-ECHAM 6
members with 12 hr diff. in initial condition
14
Coupling mechanism
The coupling is performed without flux
corrections and performed only inside REMO domain
Aldrian et al, 2005, Clim Dyn.
15
Improvement by coupling
Aldrian et al, 2005, Clim Dyn.
Java
Java
uncoupled
coupled
Southern South China Sea
Southern South China Sea
Molucca Sea
Molucca Sea
16
Local ocean aspect of the three climate regions
Aldrian and Susanto, 2003, Intl. J Climatol.
Similar results to Bony et al, 1997a,b, Lau et
al, 1997
17
Role of ocean circulation in driving the rainfall
characteristics of all three regions
Using a heat budget calculation

• 15 yr Correlations North Molucca
• Horz adv heat0.207
• Heat SST 0.394
• Horz adv SST 0.194

• 15 yr Correlations SSCS
• Horz adv heat0.613
• Heat SST 0.711
• Horz adv SST 0.842

18
ENSO episodes between rainfall and NINO3 SST

• impact starts in April
• No impacts on the peak of rainfall season in DJF
• The ocean mechanism prevents the impact during
  the wet season
• El Nino is in-phase to the dry season, thus
  worsen the case
• La Nina has no impact in the peak of the dry
  season MJJ because they are out-of-phase

19
The INSIDE Project

• An European Union Asia ProEco program 2004-2006
• The INSIDE project aims to determine the amount
  and distribution of smoke haze in Indonesia and
  the adjacent countries generated from vegetation
  and peat fires, and the related implications for
  human health (e.g. respiratory diseases) and
  climate (droughts, floods, aerosol-cloud
  interactions, CO2 release). The main goal of the
  project is to provide, optimize and apply a
  regional model tool for Indonesia. Due to the
  sparse air quality monitoring in Indonesia our
  initiative with the country-wide determination of
  ambient air quality offers guide to local
  decision makers

20
Examples of modeling result
Heil, Langmann, Aldrian, 2007 Mit. Adapt. Stra.
Global Change
Distribution of PM10 concentration
Comparison of observed and simulated particle
concentration over Kuching (northwest Borneo) and
Petaling Jaya (Malaysian Peninsula)
21
Local SST switch on and off
22
Sea air interaction over Mollucca
Example of southern sea
Observed Mollucca Sea
Previous results, Aldrian and Susanto 2003
Preliminary results
23
?SST vs ?Precip (Jan-Jun)
24
?SST vs ?Precip (Jul-Dec)
25
Annual march of SST rainfall relationship

• Changes in SST affects more on convective type of
  rainfall or the local type with also much higher
  correlation values (R).

26
The LITHMOS Project

• A DFG Germany proposed project
• Simulate local ocean with boundary forcing from
  the coupled ocean atmosphere model
• Use forcing from NCEP or ERA40
• Use observed deep ocean data from the INSTANT
  project (2004-2006) of 11 mooring locations

27
Climate trend of rainfall in IndonesiaAldrian
2006, National Publication

• Investigation from 1960 to 1998 in 63 major
  stations

28
Aldrian, 2006
29
Annual water lost potency of large islands due to
decreasing rainfall
30
Monsoon weakening
Dec
May

• Data from Brantas catchment east Java for the
  last 51 years (1955 2005)

Aldrian and Djamil 2008, Intl J Climatol
31
Extension of dryspell
Aldrian and Djamil 2008, Intl J Climatol
32
Change of ratio rain falls in wet season to
total(east Java)

• Higher risk of flood and drought

Aldrian and Djamil 2008, Intl J Climatol
33

• The leading annual eigen coefficients of PC1 and
  PC2 (black lines) along with the NINO3 aSST (grey
  lines in unit oC) for the PC1. For the PC2, the
  linear straight grey lines highlight the
  bi-decadal variability at level 0.0, 0.2 and
  -0.18 during the period 1955-1973, 1974-1988 and
  1989-2005, respectively.

34

• The leading seasonal eigen coefficients of PC1
  (black lines) after removing the annual signal
  along with the NINO3 aSST (grey lines in unit
  oC).

Aldrian and Djamil 2008, Intl J Climatol
35
Conclusions

• the rainfall climate of this region is
  predictable at monthly and seasonal scales, but
  only for a limited and specific period in
  specific regions. The predictability barrier is
  an intrinsic character of Indonesian rainfall and
  a challenge to climate modelling in the region
  because it limits model applications.
• beside monsoon, important rainfall variability
  from monthly to interannually is the coherent
  ENSO.
• The global atmospheric models produce the large
  scale precipitation characteristics well, but the
  regional model shows a regional phenomenon
  obscured within global models through better
  orography.
• A coupled regional atmosphere/ocean model shows
  improved dynamics through the ocean and the
  atmosphere sea-air interaction and feedback.
• The ocean regulates the SST with persistence time
  lag at the ocean surface. The horizontal
  advection fluxes changes the heat content of the
  water column, which eventually changes the
  surface condition (SST). The SST regulates the
  local precipitation through a specific SST
  rainfall relationship.
• Decreasing climatic trend of rainfall due to
  global climate change with much decrease of
  rainfall in dry period than increase of rainfall
  in wet period.

36
Other climate related researches

• May 2005-September 2006. The Brantas Catchment
  Water and Carbon Cycle, Funding Southeast Asia
  Regional Committee START, Project No
  94/01/CW-001. Reference Prof. Chou Han Liu,
• July 2006 June 2009. LITHMOS (Long-term
  Indonesian THroughflow MOdel Simulation).
  Funding Deutsche Forschung Gemeinschaft / Bundes
  Ministrium fur Zusammenarbeit (DFG/BMZ) Project
  No DA 773/3-1 and FUNDING from the Indonesian
  Govt. Incentive Research 2008-2009, Reference
  Dr. Peter Damm,
• May 2006 October 2007 Nutrient, sediment and
  water fluxes of Java river basins, Funding
  Southeast Asia Regional Committee START, Project
  No 95/01/CW-002. Reference Prof. Cheng-Tung
  Arthur Chen, Prof. Hidayat Pawitan and FUNDING
  from the International Foundation of Science 2008
• May 2007 October 2008 Water and Carbon Cycle
  under intensive Human and Peat influenced of Musi
  River, Funding Southeast Asia Regional Committee
  START, Project No 96/01/CW-003. Reference Prof.
  Cheng-Tung Arthur Chen,
• January 2007-December 2007. Paleo climate study
  of the Jakarta Bay through coral proxy Funding
  Competitive Program of the Indonesian Institute
  of Science (LIPI) DAS JABOPUNCUR, Project No
  134.C/IPK.3/KS/2007. Reference Dr. Yudawati
  Cahyarini,
• January 2008-December 2009. Study on the use of
  weather radar, satellite and numerical model for
  flood risk management of Jakarta Metropolitan
  area, FUNDING Indonesian government incentive
  research 2008-2009.

37
Thank You very much for your kind attention