Conference PaperPDF Available

The CSIRO-Mk3.6.0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication


Abstract and Figures

The participation of the CSIRO-Mk3.6.0 Atmosphere Ocean Global Climate Model (AOGCM) in the Coupled Model Intercomparison Project Phase 5 (CMIP5) is a joint initiative between the Queensland Climate Change Centre of Excellence and the Commonwealth Scientific and Industrial Research Organisation (CSIRO). It now has approximately 10 research and support scientists working on this project which first began in 2009. This on-going project consists of the following four main components: • A model design and testing period to ensure that the model had acceptable configuration for participation in CMIP5, in particular, exhibiting a realistic present-day climate and a stable pre-industrial climate;
Content may be subject to copyright.
The CSIRO-Mk3.6.0 Atmosphere-Ocean GCM:
participation in CMIP5 and data publication
M.A. Colliera, S.J. Jeffreyb, L.D. Rotstayna, K.K-H. Wongb, S. M. Dravitzkia ,C. Moesenederc,
C. Hamalainenb, J.I. Syktusb, R. Suppiaha, J. Antonyd, A. El Zeind and M. Atifd
a The Centre for Australian Weather and Climate Research, CSIRO Marine and Atmospheric Research,
Aspendale, Victoria
b The Queensland Climate Change Centre of Excellence, Ecosciences Precinct, Dutton Park, Queensland
cCSIRO Marine and Atmospheric Research, Ecosciences Precinct , Dutton Park, Queensland
dNational Computation Infrastructure National Facility, Australian National University, Australian Capital
Abstract: The participation of the CSIRO-Mk3.6.0 Atmosphere Ocean Global Climate Model (AOGCM)
in the Coupled Model Intercomparison Project Phase 5 (CMIP5) is a joint initiative between the Queensland
Climate Change Centre of Excellence and the Commonwealth Scientific and Industrial Research
Organisation (CSIRO). It now has approximately 10 research and support scientists working on this project
which first began in 2009. This on-going project consists of the following four main components:
A model design and testing period to ensure that the model had acceptable configuration for
participation in CMIP5, in particular, exhibiting a realistic present-day climate and a stable pre-
industrial climate;
A model integration phase where CMIP5 experiments were performed. These were to include the
so-called “core” experiments plus a number of “tier1” and “tier2” experiments, which will constitute
a significant submission to CMIP5 and to address local climate modelling needs and applications;
Post-processing of the raw CSIRO-Mk3.6.0 model output into internationally recognised and
standardized CMIP5 form; and
Quality control and publication phase of the CSIRO-Mk3.6.0 data to ensure entry into the Earth
System Grid (ESG) Federation, allowing it to be disseminated to the CMIP5 international
In this paper the four phases of this climate modelling project will be discussed in detail. The main emphasis
is to make potentially interested researchers aware of the CSIRO-Mk3.6.0 climate model submission and to
elucidate the range and features of the datasets that are now available. The CMIP5 datasets are being hosted
on the ESG which consists of international data nodes and gateways, including Australia’s own node hosted
by the National Computing Infrastructure (NCI) National Facility in Canberra. A key outcome of our efforts
is the generation of over 150, mostly high priority, uniquely defined parameters from the list of requested
model output to understand climate processes and also produce new climate change projection data for
impact assessment. Some preliminary results of the CSIRO-Mk3.6.0 model are presented to illustrate the
usefulness of this dataset in this research area.
Keywords: CMIP5, AR5, AOGCM, climate change simulations, Earth System Grid
19th International Congress on Modelling and Simulation, Perth, Australia, 12–16 December 2011
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
The partnership between the Commonwealth Scientific and Industrial Research Organisation (CSIRO) and
Queensland Climate Change Centre of Excellence (QCCCE) in the Coupled Model Intercomparison Project
Phase 5 (CMIP5) has quite possibly delivered the largest and scientifically most comprehensive set of
Atmosphere-Ocean Global Climate Model data ever generated in Australia. Although there is a need to make
these datasets available to climate analysts around the world, there is a particular need in Australia for a
comprehensive set of climate experiments. This will facilitate research into the important drivers of Australian
climate. CSIRO has participated in the CMIP modelling activity before (Collier et al., 2007) however the data
volume is an order of magnitude greater than the previous model intercomparison (Collier et al., 2011a,b).
CMIP5 is an internationally coordinated effort to use state-of-the-art Global Climate Models (GCMs) and Earth
System Models (GCMs which include interactive carbon and/or ocean biogeochemistry) to perform a set of pre-
defined experiments (Taylor et al., 2011). Scientific publications arising from analysis of CMIP5 data will be
used to assess climate change science, most directly through the publication of the Intergovernmental Panel on
Climate Change (IPCC) Fifth Assessment Report (AR5) due to be published in 2013. CMIP5 has taken datasets
from a range of internationally recognised climate models and has built an archive that can be readily analysed
because all datasets have a consistent format. The strict formatting requirements also necessitated an
unprecedented approach to quality control (see Sec. 5 for details). To end-users, the CMIP5 data archive will
provide an efficient means for locating and obtaining datasets for local analysis. CMIP5 datasets are being made
publicly available via the Earth System Grid (ESG) federation of services. Users will identify the desired
dataset(s) through a series of ESG gateways and will then be directed to the appropriate data node to download
the desired dataset(s) to their local machine.
The CSIRO-Mk3.6.0 model, hereafter called Mk3.6, is an upgrade from the CSIRO-Mk3.5 GCM (Gordon et al.,
2010). Details of the model are given by Rotstayn et al. (2010). The atmospheric component has a horizontal
resolution of approximately 1.9°x1.9° and every atmospheric grid-point is coupled to two ocean grid-points.
This enhanced north-south resolution in the ocean component is expected to increase the capacity for the ocean
to simulate important tropical and extra-tropical seasonal interactions. The atmosphere has 18 vertical levels
whereas the ocean has 30 levels with most found in the upper 1500m. By far the most important improvement
of the Mk3.6 model from its predecessor is the inclusion of an interactive aerosol scheme that also required an
update to the radiation scheme used in the model (Rotstayn et al., 2010). This allows for the investigation of the
impact of a number of aerosol agents on climate. For example, a recent study by Rotstayn et al. (2011a)
investigated the impact of mineral dust on Australian rainfall by turning it on and off in two experiments. The
study found that an accurate simulation of the El Niño-Southern oscillation (ENSO)-rainfall relationship over
Australia might require realistic representation of processes associated with sources and deposition of Australian
Figure 1. Global average surface air
temperature (°C) for the CSIRO-Mk3.6.0 pre-
industrial control experiment. The dashed line is
the 500 year average.
Figure 2. Global average volume weighted a)
ocean potential temperature (°C) and b) salinity
(psu-34.0) for the CSIRO-Mk3.6.0 pre-
industrial control experiment.
To have confidence in a climate model’s ability to realistically simulate present and future climate conditions it
is necessary for it to be able to respond in a satisfactory manner when driven by pre-industrial (year 1850)
forcings. An important indicator of this is the stability of the model solution which includes negligible drift in
important climate indices and one devoid of irregular behavior. Figure 1 shows globally averaged annual
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
surface temperature from the pre-industrial control experiment, which appears to be stable during the 500 year
long experiment. Figure 2 shows the ocean potential temperature and salinity (practical salinity units). The
ocean temperature drift in the Mk3.6 is 0.02°C/century, which is comparable to other coupled climate models.
The salinity drift is also and indicates the global precipitation and evaporation are not perfectly balanced in the
model, at least on these time-scales.
In this section we describe the computing facilities that were used for the integration of the Mk3.6 AOGCM for
CMIP5. In addition the list of experiments will be presented with essential details on their characteristics,
including their name, the number of ensembles and the output model years.
4.1. QCCCE Computing Facility
The complete set of Mk3.6 experiments was run on the Queensland Government Department of Environment
and Resource Management’s High Performance Computing facilities. Original experimentation began in
January 2010 and most of the key experimentation finished in July 2011, however, the long model integrations
out to the year 2300 are expected to be finished in late 2011. Although the computing resources were adequate
for this project the model output needed to be transferred to the National Computing Infrastructure (NCI)
National Facility (NF) in Canberra for data hosting.
The CMIP5 experiments conducted with the Mk3.6 climate model are listed in Table 1. Most of the experiments
were performed using a fully coupled (AOGCM) whereas some with an atmosphere/land/sea-ice only (AGCM).
See and http://cmip- for details of experimental design and on standard
naming conventions.
Table 1. CSIRO-Mk3.6.0 CMIP5 experiments. See text for details. Notes: ensemble members 1-3 are
extended to 2300; †† ensemble members 2-12 are 5 years in length consistent with the CMIP5
specification; and ††† experiment commenced in 1950 as ozone changes prior to 1950 were considered
Experiment CMIP5 Experiment
Type Ensemble size Years
piControl 3.1 AOGCM 1 1-500
historical 3.2 AOGCM 10 1850-2005
amip 3.3 AGCM 10 1979-2009
midHolocene 3.4 AOGCM 1 1-100
rcp45 4.1 AOGCM 10 2006-2100
rcp85 4.2 AOGCM 10 2006-2100
rcp26 4.3 AOGCM 10 2006-2100
rcp60 4.4 AOGCM 10 2006-2100
1pctCO2 6.1 AOGCM 1 1-140
sstClim 6.2a AGCM 1 30
sstClim4xCO2 6.2b AGCM 1 30
abrupt4xCO2 6.3 AOGCM 12 1-150††
sstClimAerosol 6.4a AGCM 1 1-30
sstClimSulfate 6.4b AGCM 1 1-30
historicalNat 7.1 AOGCM 10 1850-2012
historicalGHG 7.2 AOGCM 10 1850-2005
historicalAnt 7.3a AOGCM 10 1850-2005
historicalNoOz 7.3b AOGCM 10 1950-2012†††
historicalNoAA 7.3c AOGCM 10 1850-2005
historicalAA 7.3d AOGCM 10 1850-2012
historicalAntNoAA1 7.3e AOGCM 10 1850-2012
The model development and experimentation was a significant challenge: Mk3.6 is the culmination of over 30
years of model development (Smith, 2007). Once the experimentation was complete, the post-processing and
publishing cycles also required a substantial amount of work and data processing. Post-processing of model
1 This experiment was designed to isolate the effect of Asian aerosols, in the manner of Rotstayn et al. (2007).
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
output was done using the Coupled Model Output Rewriter (CMOR,, and
publishing of model data was partially automated by the Earth System Grid data hosting infrastructure.
CMOR is a library that can be used to reformat datasets to a standard prescribed in user-defined tables. By
adopting CMOR, the task of reformatting the raw model output to CMIP5 specifications was reduced to writing
customised software to load the raw datasets and perform any required derivations or modifications of data. The
model data was converted to CMIP5 form using CMOR format-specification tables provided by CMIP5. In
addition to reducing the complexity of the post-processing task, the use of CMOR is also expected to improve
the quality of the data because CMOR: (i) performs some rudimentary error checking; and (ii) automatically
formats the metadata to CMIP5 standards.
The in-house software package can become extremely complicated and sophisticated in itself, as it has to
consider technical issues associated with the raw model output and supply the necessary objects to CMOR.
Necessary inputs can be simple text strings like the institution name but could also involve complex
calculations, for example the derivations of parameters or interpolation from model hybrid coordinates to
standard pressure levels.
This section gives details of the Quality Control (QC) approach taken to ensure that the Mk3.6 model output
satisfy the CMIP5 standard. An explanation of how the final submitted data were published is also provided.
6.1. The QC and QCWrapper utilities
One of the most significant shortcomings of the previous activity CMIP3 was the inadequate level of QC
conducted on datasets. CMIP5 has a range of QC Levels (QCLs) which are performed at different stages of the
post-processing and publishing cycle. By adopting the CMOR interface QCL1 and QCL2 standards are
essentially achieved (see for background information). For more comprehensive
checking, the QC tool was used to check all Mk3.6 datasets that were submitted. QC uses a wrapper to impose
project-specific requirements; in this case the CMIP5 wrapper was used. When examining a file, QC checks the
time coordinate, metadata and data block. The data are scanned to detect values that are missing or replicated,
and some statistical properties, such as the global maximum, minimum, mean and standard deviation are
computed. While the information from the QC tool is very useful, it is still nevertheless at the discretion of the
modeling centre to act on any warnings or errors provided. One extremely useful output from the QC tool is a
NetCDF file containing the global mean and standard deviation for each time slice in the input file that was
examined. Plotting the mean and standard deviation can be useful in detecting gross errors in the model data
and/or processing system. While there may be hundreds of plots for each experiment, these can be scanned
through quite quickly. Final QCL3 checks will eventually be performed by the ESG community on the archived
datasets allowing the allocation of a Digital Objective Identifier (DOI) in essence giving the datasets persistence
and citable credentials in the digital environment.
In this section we will present some results based on annual average conditions, particularly focusing on near
(2m) surface air temperature and rainfall. In the future we expect to expand this work focusing on seasonally
based temperature, mean sea-level pressure and rainfall projections over Australia.
By the end of August 2011 the processed output from the CSIRO-Mk3.6.0 model for a number of key CMIP5
experiments had been published on the NCI ESG gateway. It is expected that the research community both in
Australia and abroad will undertake extensive analysis of these datasets. Prior to peer-review publications,
preliminary results based on the raw model output have been published elsewhere (see for example, Syktus et al.
In this section we present some results based on annual average conditions, particularly focusing on near (2m)
surface air temperature and rainfall.
The model simulated global average near-surface
air temperature for the period 1850-2055 and
observed (Brohan et al., 2006) data from 1850 to
2010 are shown in Figure 3. The data are presented
as anomalies relative to the 1850-1879 base period.
The results for four historical experiments are
presented: (i) the historical run with all forcings
extended to 2100 by using forcing data from
Representative Concentration Pathway (RCP) 4.5
(HIST/RCP4.5); (ii) natural forcings only (NAT);
(iii) greenhouse gas forcings only (GHG); and (iv)
anthropogenic aerosol forcings only (AA). The
NAT, GHG and AA experiments are driven by the
observed values for the relevant forcing (natural,
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
greenhouse gases and anthropogenic aerosols,
respectively), with all other forcings held constant
at pre-industrial (1850) levels. All model
experiment data are based on a 5 member ensemble
average. The HIST experiment shows the best
agreement with the observations, as also seen in
earlier simulations that include aerosol forcing.
Figure 3. Global average surface air
temperature (°C) for four CSIRO-Mk3.6.0
experiments and HadCRUT observations. Filled
areas show the range based on the 5 member
ensemble and solid lines are ensemble means.
Table 2. Summary of statistics for near surface
air temperature (Tsc) and annual average
precipitation (Pr) for the period 1980-2005 for
the all-forcings (HIST) experiment 5 member
ensemble. Average (ave), standard-devation (sd),
root-mean-square (rms) error and pattern
correlation (corr) are shown. Minimum and
maximum ensemble values are shown by
subscripts min and max respectively.
Observations have been interpolated onto the
model grid for calculating these statistics.
AWAP observational values are shown in
avemin avemax sd rms corr
Tsc 21.02
20.96 21.13 0.066 1.52 0.95
Pr 1.37
1.31 1.40 0.036 0.58 0.78
The model simulated and observed AWAP (Australian Water Availability Project, Jones et al., 2009) near-
surface air temperature and precipitation for Australia are shown in Figure 4. The 5 member ensemble mean for
the period 1980-2005 is presented, with the ensemble standard deviation indicated by hatching. Regions
exhibiting a relatively high standard deviation indicate a wide range in the ensemble members, indicating the
potential for more uncertainty in the ensemble mean due to different forcings. A comparison of spatial patterns
of simulated temperature for the HIST experiment (Figure 4a) and observations (Figure 4e) indicates the Mk3.6
model reproduces the observed pattern, although the model underestimates the mean in the south and
continental interior. The historical experiment driven only by natural forcings (NAT) (Figure 4b) shows slightly
lower temperatures compared to the all forcings experiment (Figure 4a), while the experiment driven only by
greenhouse gases (GHG) (Figure 4c) overestimates the temperature over the continent. In contrast, the
experiment driven only by anthropogenic aerosols (AA) (Figure 4d) slightly underestimates the observed
pattern, consistent with the net cooling effect expected of such aerosols. Maps showing the differences between
the results of the various attribution experiments would enable greater differentiation between the impacts of the
various drivers, and will be the topic of further investigation. Standard deviations based on ensemble members
are higher over marginal areas of southeast and central northwest.
The model simulated precipitation is shown by panels f-j of Figure 4 indicating good agreement with the
observed spatial pattern, although there appears to be a dry bias, particularly in the south-west and south-east. It
should be noted that the annual average rainfall does not reflect the important characteristics of the seasonal
rainfall distribution and therefore an analysis of the seasonal rainfall distribution will be required for a better
assessment of model skill in simulating the Australian rainfall.
An analysis of attribution experiments can provide important assessment potential roles of various climate
forcing factors that affect Australian climate. A recent study by Rotstayn et al (2011b) provides an insightful
example of the technique.
The model performance in the Australian region for the all-forcings HIST experiment is summarised in Table 2.
The ensemble mean near surface air temperature is 21.03 °C for the period 1980-2005, compared to an observed
value of 22.04 °C. The root mean square error is 1.52 °C and the pattern correlation is 0.95. The ensemble mean
annual average precipitation is 1.37 mm/day for the period 1980-2005, compared to an observed value of 1.30
mm/day. The root mean square error is 0.58 mm/day and the pattern correlation is 0.78. The statistics for both
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
near surface air temperature and precipitation indicate sound model performance using the demerit point system
of Suppiah et al. (2007). Standard deviations based on the 5-member ensemble are higher over central and
eastern Australia.
Figure 4. Annual average near surface air temperature (panels a-e, °C) and precipitation (panels f-j,
mm/day) simulated by the CSIRO-Mk3.6.0 and observations (AWAP). The hatching indicates the 5
member ensemble standard deviation (°C and mm/day, respectively). HIST, NAT, GHG, AA refer to
experiments historical, historicalNat, historicalGHG and historicalAA from Table 1.
Collier et al., The CSIRO Mk 3-6-0 Atmosphere-Ocean GCM: participation in CMIP5 and data publication
In this paper we have described the experimentation, post-processing, quality control and publishing phases
involved in preparing the Mk3.6 datasets for submission to the CMIP5 data archive. The most novel aspects of
the submission are the relatively large ensemble sizes used in the experiments, and the range of historical
experiments undertaken. The attribution experiments will provide a rich dataset for elucidating the key drivers
of change in Australia’s climate. It is envisaged that the submission will be used in many climate change
detection and attribution studies that will be used to prepare the IPCC 5th Assessment Report, due for release in
2013. Moreover, the datasets will be of great interest to the Australian scientific community well beyond AR5,
providing the agencies that supported this work with a significant return on investment. Our early analysis
suggests that the Mk3.6 model soundly simulates present day screen temperature and precipitation over
Australia, lending credence to the future projections generated by the model.
Brohan, P., J. J.Kennedy, I. Harris, S. F. B. Tett, and P. D. Jones (2006), Uncertainty estimates in regional and
global observed temperature changes: A new data set from 1850, J. Geophys. Res., 111, D12106,
Collier, M.A., Dix, M.R. and A.C. Hirst (2007). CSIRO Mk3 Climate System Model and Meeting the Strict
IPCC AR4 Data Requirements. MODSIM08 Extended Abstract, Christchurch, New Zealand.
Collier, M.A., Jeffrey, S., Dix, M.R., Hirst, A.C and L.D. Rotstayn (2011a). Dealing with and contributing to
the CMIP5 data ‘tsunami’ and beyond from an Australian perspective. Greenhouse 2001, Cairns, 4-18th April
Collier, M.A., Jeffrey, S. and L.D. Rotstayn (2011b). The latest Australian CMIP climate model submission.
BAMOS, October edition, in press.
Gordon, H.B. and co-authors (2010). The CSIRO Mk3.5 Climate Model. CAWCR Technical Report No. 21,
Jones, D. A., Wang, W., and Fawcett, R. (2009) High-quality spatial climate data-sets for Australia, Aust.
Meteorol. Oceanogr. J., 58, 233–248.
Rotstayn, L.D., Collier, M.A., Dix, M.R., Feng, Y., Gordon, H.B., O’Farrell, S.P., Smith, I.N. and J.I. Syktus
(2010). Improved Simulation of Australian Climate and ENSO-related rainfall variability in a global climate
model with interactive aerosol treatment. Int. J. Climatol. 30: 1067-1088. doi: 10.1002/joc.1952.
Rotstayn, L.D., Collier, M.A., Mitchell, R.M., Qin, Y., S.K. Campbell and S. M. Dravitzki (2011a). Simulated
enhancement of ENSO-related rainfall variability due to Australian dust. Atmos. Chem. Phys., 11, 6575–6592,
doi: 10.5194/acp-11-6575-2011.
Rotstayn, L.D., Jeffrey, S.J., Syktus, J.I., Collier, M.A., Dravitzki, S.M., Hirst, A.C. and K.K-H. Wong (2011b).
Have anthropogenic aerosols delayed greenhouse gas-induced changes in Indo-Pacific regional circulation and
rainfall? Atmospheric Science Letters, submitted.
Smith, I. (2007). Global climate modelling within CSIRO: 1981 to 2006, Aust. Meteorol. Mag., 56, 153-166.
Suppiah, R., Hennessy, K.J., Whetton, P.H., McInnes, K., Macadam, I, Bathols, J., Ricketts, J. and C.M. Page
(2007). Australian climate change projections derived from simulations performed for the IPCC 4th Assessment
Report, Aust. Met. Mag., 56, 131-152.
Syktus, J., Wong, K.K-H., Rotstayn, L.D., Jeffrey, S., Zhang, H., Toombs, N.R. and M.A. Collier (2011).
Australia’s hotter and drier future: Climate change projections using CMIP5 experimental design and the
CSIRO Mk3.6 climate model. Greenhouse 2011, Cairns, 4-18th April 2011.
Taylor, K.E., Stouffer, R.J. and G.A. Meehl (2011). A Summary of the CMIP5 Experiment Design. Available in
electronic form:
... An advantage of direct simulation is the detection algorithm, and no additional assumptions are required [5] when compared to the other two methods. For the direct simulation of typhoon in the WNP, both fine (~20-50 km) and coarse (~100-300 km) resolution climate models have been used [6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. They can have relatively good representations of the large-scale processes that affect typhoon characteristics such as genesis locations and tracks [21,22], though typhoon intensities are poorly resolved in coarse resolution models. ...
Full-text available
This paper presents a development of a statistical model of typhoon genesis, tracks based on kernel density estimation and a generalized additive model (GAM). Modeling of typhoon activity is ultimately beneficial to the people living in coastal zones, insurance/re-insurance companies, policy, planning and decision departments. A 50-year record (1972-2021) of typhoon track observations from the International Best Track Archive Climate Stewardship have been used to observe the distribution of typhoon genesis by kernel density estimation. The tracks are simulated through the development of a GAM. It reproduces the observation well. A distance calculation approach between observed and simulated tracks' landfall have been used to validate the model and the model shows a very good skill (approximately 75%).
... Carbon dioxide (CO 2 ) concentrations (ppm) comparison for SRES scenarios B1, A1B, A2 and A1FI, and RCP2.6, RCP4.5, RCP6, and RCP8.5(Collier et al. 2011). Base period of 20 years from 1981 to 2000 were taken for these estimations/predictions. (Source RCP database version 2.0.5) ...
Agricultural production is low in many parts of the world which is subjected to limited adaptability toward adverse events. Climate change (CC) is estimated to decrease productivity on even lower levels and has made production more inconsistent. Many countries all over the world have planned to accept climate smart agriculture (CSA) approach to make improvements in agriculture. The CSA refers to a combined set of technologies and practices that simultaneously improve farm productivity as well as monetary return, enhance adaptability to CC, and minimize the emissions of greenhouse gas (GHG). The concept of CSA is gaining significance at national and international levels to cope with future challenges of the agricultural-related plannings. The CSA is a notion that calls for the combination of the need of adaptability to CC and strategies for mitigation in agriculture to ensure food security. For instance, strategies like the use of renewable energy for agriculture, i.e., pyrolysis units, solar panels, windmills, as well as water pumps are very vital for food production. The concept of CSA comprehends three major stakes (productivity, adaptation, and mitigation), but the literature has not fully addressed them in an integrated way. Adaptation and productivity were of prime importance for poor and under developing countries, while mitigation was mainly addressed in developed countries. The utmost challenge for policymakers and stakeholders to operationalize CSA is the identification, assessment (cost-benefit ratio), as well as subsequent ordering of CSA options and portfolios for investment. The aims of CSA are to sustainably increase farmer’s resilience, improve agricultural productivity, achieve food security and sustainable development goals in addition to reduce emissions of GHG.
... For future climate scenarios, we selected three global General Circulation Models (GCMs) from the Coupled Model Intercomparison Project 5 (CMIP5) namely, MRI-CGCM3 (Yukimoto et al., 2012), CISRO-MK3.6 (Collier et al., 2011) and NIMR-HADGEM2-AO (Baek et al., 2013). Each GCM was tested under three Representative Concentration Pathways (RCP 4.5, RCP 6.0 and RCP 8.5) for 2050 and 2070 as described in the IPCC fifth assessment report (IPCC, 2013). ...
Identifying the distributional range and climatically suitable habitat of a species are crucial for developing an effective conservation strategy. This study aims to simulate the current distribution of climatically suitable habitat for a threatened medicinal tree, Oroxylum indicum (L.) Kurz in India and predict its possible range shifts in future climate scenario. Combined with 19 bioclimatic variables from WorldClim, we used Maxent to predict the potential suitable areas for species in current climatic condition and future climate under three Representative Concentration Pathway (RCP) scenarios by integrating three General Circulation Models (GCMs) for 2050 and 2070. Furthermore, we performed change analysis to identify the precise difference between the current and future distribution of suitable areas of the species for delineating habitat range expansion (gain), habitat contraction (loss) and stable habitats. The model predicted ∼25.47% of the geographical area in India is presently climatically suitable for the species. The study revealed that annual precipitation, precipitation of wettest and warmest quarter, precipitation of wettest month, maximum temperature of warmest month and mean temperature of warmest quarter significantly influence the distribution of the studied species. The net habitat loss under three RCP scenarios (RCP 4.5, 6.0 and 8.5) was estimated to range from 10.51% to 17.31% for the year 2050, which could further increase 16.63% to 20.40% by the year 2070. The finding indicates that rise in temperature could significantly reduce the potential habitat of O. indicum and hence suitable measures need to be taken to conserve this important medicinal plant.
Full-text available
Climatic design conditions are widely used by the building community as environmental parameters informing the size and energy requirements for heating, ventilation and air conditioning systems, along with other building design characteristics. Climatic design conditions are calculated by the American Society of Heating, Refrigerating and Air-conditioning Engineers using historical climate data. Our work advances methods for projecting future climate design conditions based on data from global climate models. These models do not typically archive the hourly data required for climate design condition calculations, and they often exhibit large biases in extreme conditions, daily minimum temperatures and daily maximum temperatures needed for climatic design conditions. We present a method for rescaling historical hourly data under future climatic states to estimate the impact of climate change on future building climatic design conditions. This rescaling method is then used to calculate future climatic design conditions in Madison, Wisconsin, throughout the 21st century for two future greenhouse gas emissions scenarios. The results are consistent with a warming climate and show increases in heating, cooling, humidification and dehumidification design conditions, suggesting less extreme cold conditions and more extreme hot and humid conditions in Madison. The design conditions used for estimating energy demand, degree days, show that under a business-as-usual scenario, by the mid-century, building heating and cooling in Madison (climate zone 5A) will be similar to the current heating demand in Chicago, IL (climate zone 5A) and cooling demand in Baltimore, MD (climate zone 4A); by the late-century, building heating and cooling in Madison will resemble the current heating demand in St Louis, MO (climate zone 4A) and cooling demand in Augusta, GA (climate zone 3A). Given the rapid pace of climate change in the 21st century, our work suggests that historical design conditions may become obsolete during even the initial stages of a building’s expected life span. Changes in climatic design conditions in Madison highlight the importance of considering future climatic changes in building design to ensure that buildings built today meet the performance needs of the future.
Full-text available
The global mean surface temperature cooled slightly in the mid-twentieth century despite a continuous increase in greenhouse gas concentrations. The cooling was strongest in the Northern Hemisphere mid-latitudes, while the Southern Hemisphere mid-latitudes experienced moderate warming. This apparent contradiction is often attributed to internal multi-decadal variability originating from Pacific and Atlantic ocean-atmosphere interactions. Given the rapid increase of industrial activities in North America and Europe during that period, it is also plausible that anthropogenic aerosol (AA) emissions (as an external forcing) contributed to the stronger Northern Hemisphere cooling. This paper aims to quantify the contributions of AA and decadal variability to the 1948–1978 cooling. We analyzed the latitudinal temperature trend asymmetry in 60° S–60° N throughout the troposphere, using multiple reanalysis datasets and the Coupled Model Intercomparison Project phase 5 (CMIP5) multi-model ensemble that bears significant similarity with the observed patterns. We show that both AA increase and the North Atlantic Variability Index (NAVI) transition into its negative phase are the major contributors to the latitudinal asymmetry of cooling. At the surface level, based on the horizontal pattern correlation method, AA and NAVI have similar contribution fractions (20 vs. 16%), but the contribution fraction of AA is much larger at 500 hPa (55 vs. 8%). Attributions based on vertical pattern correlation and latitudinal gradient show consistent results. Natural forcings (NAT) also contribute to the cooling asymmetry during mid-20C, but with a much smaller impact compared to AA and NAVI. Therefore, we argue that previous studies that mostly focused on surface variables may have underestimated the role of AA in the mid-twentieth-century climate change. The study suggests that the three-dimensional thermal structure and atmospheric circulation change should be closely examined in future climate attribution analysis.
This study analyzes projected heat extremes over the Middle-East–North Africa (MENA) region until the end of the twenty-first century with a number of temperature indices based on absolute values and thresholds to describe hot conditions. We use model projected daily near-surface air (2-m) temperature ( $$T_\mathrm{{max}}$$ T max and $$T_\mathrm{{min}}$$ T min ) to derive the indices for the period 1980–2100. The data were taken from 18 CMIP5 models combining historical (1980–2005) and scenario runs (2006–2100 under RCP2.6, RCP4.5, and RCP8.5 pathways). Results show a domain-wide projected warming for all emission scenarios. Our findings for a business-as-usual pathway indicate excessive warming of more than 8 $$^\circ $$ ∘ C in the northern part of the domain (south Europe) for the annual warmest day (TXx) and night (TNx). In the hottest parts of the domain record high temperatures reached 50 $$^\circ $$ ∘ C in the recent past, which could increase to at least 56 $$^\circ $$ ∘ C by the end of the century, while temperatures over 50 $$^\circ $$ ∘ C are expected to occur in a large part of the MENA region. A significant increase is projected in the number of hot days (TX $$>40^\circ $$ > 40 ∘ C) and nights (TN $$>30 ^\circ $$ > 30 ∘ C) all over the region. For the period of 2071–2100 excessive hot days and nights will become the normal during summer in large parts of the MENA with some locations expected to exceed 180 and 100 days, respectively. Calculations of the corresponding heat index suggest that several areas across the MENA region may reach temperature levels critical for human survival.
Full-text available
The largest explosive volcanic eruption of the Common Era in terms of estimated sulphur yield to the stratosphere was identified in glaciochemical records 40 years ago, and dates to the mid-thirteenth century. Despite eventual attribution to the Samalas (Rinjani) volcano in Indonesia, the eruption date remains uncertain, and the climate response only partially understood. Seeking a more global perspective on summer surface temperature and hydroclimate change following the eruption, we present an analysis of 249 tree-ring chronologies spanning the thirteenth century and representing all continents except Antarctica. Of the 170 predominantly temperature sensitive high-frequency chronologies, the earliest hints of boreal summer cooling are the growth depressions found at sites in the western US and Canada in 1257 CE. If this response is a result of Samalas, it would be consistent with an eruption window of circa May–July 1257 CE. More widespread summer cooling across the mid-latitudes of North America and Eurasia is pronounced in 1258, while records from Scandinavia and Siberia reveal peak cooling in 1259. In contrast to the marked post-Samalas temperature response at high-elevation sites in the Northern Hemisphere, no strong hydroclimatic anomalies emerge from the 79 precipitation-sensitive chronologies. Although our findings remain spatially biased towards the western US and central Europe, and growth-climate response patterns are not always dominated by a single meteorological factor, this study offers a global proxy framework for the evaluation of paleoclimate model simulations.
Full-text available
Natural resource managers need better estimates of water storage and supply in forested landscapes. These estimates would aid planning for management activities that maintain and enhance forest health and productivity and help prepare forested landscapes for a changing climate. In particular, low soil moisture in combination with high evaporative demands can induce significant stresses on forests, increasing vulnerability to attacks of insect and disease, as well as increasing wildfire risk. Although high-resolution soils data exist for much of the Pacific Northwest, regional-scale datasets that identify forested areas potentially vulnerable to soil moisture-related drought do not exist. In this study, we used readily available spatial datasets depicting available water supply, soil depth, and evapotranspiration to model the likelihood that soils experience prolonged summer drying.
Climatic changes are associated with fluctuations spanning over a period of three decades as a classic period of computing weather trends all around the world which, by studies till now, was proved to be harmful for life on earth. Natural processes going on in this earth were observed to be impacted significantly by these variations in our climate that are the result of anthropogenic activities. Rapid growth in population demands more resources for their survival that includes the basic amenities of livelihood, i.e., nutrition, energy, and housing. Limited resources in combination with the risk of climatic changes are in fact a big problem that must be solved before it results in nonreversible damage. Modeling is the advanced approach to study climate change. Right after the Second World War, predominantly in the USA, by the end of the 1960s, representatives were being presented with the model’s findings, which strongly supported the concept that the persistent intensification in greenhouse gas (GHG) emissions caused by human activities have completely changed the overall impact of global climate. With the passage of time, more advancement in modeling was observed; first of all, conceptual models were formed; those were replaced by analog models and then energy balance models were introduced by researchers. In agricultural systems, modeling as an essential tool is accomplished by scientists from different disciplines that has contributed for six decades in this field. Models have been used in ecosystem studies, hydrology, climate, crops, livestock and Hadley Climate model version 3 (HadCM3) is recently commonly used and several other Global climate models (GCMs) are in practice apart from statistical models like Statistical Downscaling Model (SDSM) are prominent among others for analytical climatic data studies. In order to study the climate changes; different climate projection scenarios have been made on the basis of previously provided data, i.e., rainfall, temperature, carbon dioxide and GHG emissions, and other components. On the basis of these scenarios, future predictions are likely to be more realistic and hopefully helpful for addressing the changing climatic situations across the globe and proactively devising mitigation practices to save the masses.
Full-text available
A spatio-temporal analysis of representative concentration pathways (RCPs)-based projections of surface air temperatures under different combinations of global circulation models (GCMs) and RCPs for Satluj River Basin has been presented. The projections of temperature anomalies have been obtained for several meteorological stations in the basin using 16 different combinations of emission scenarios and global climate models (GCMs). For each combination, the projections of temperature anomalies have been analysed for two future time periods centred at 2030 and 2050. The results of the analysis conducted herein clearly indicate that the temperature will rise for all the future time scales with the maximum increase being projected under RCP8.5 compared to the baseline 1986–2005. However, there is large inter-model variability in the projections of temperature anomalies under different RCPs. Under RCP 8.5, the average temperature in Satluj basin is projected to rise by around 4°C by mid of the twenty-first century. The projections of temperature anomalies analysed herein could be potentially used for the evaluation of hydrological impacts of climate change in Satluj River Basin—a key basin in the Himalayan region from the view point of global warming.
Full-text available
Australian dust emissions are highly episodic, and this may increase the importance of Australian dust as a cli-mate feedback agent. We compare two 160-year coupled atmosphere-ocean simulations of modern-day climate using the CSIRO Mark 3.6 global climate model (GCM). The first run (DUST) includes an interactive treatment of mineral dust and its direct radiative effects. The second run (NODUST) is otherwise identical, but has the Australian dust source set to zero. We focus on the austral spring season, when the correlation between rainfall and the El Niño Southern Os-cillation (ENSO) is strongest over Australia. The ENSO-rainfall relationship over eastern Australia is stronger in the DUST run: dry (El Niño) years tend to be drier, and wet (La Niña) years wetter. The amplification of ENSO-related rainfall variability over eastern Australia represents an im-provement relative to observations. The effect is driven by ENSO-related anomalies in radiative forcing by Australian dust over the south-west Pacific Ocean; these anomalies in-crease (decrease) surface evaporation in La Niña (El Niño) years. Some of this moisture is advected towards eastern Australia, where increased (decreased) moisture convergence in La Niña (El Niño) years increases the amplitude of ENSO-related rainfall variability. The modulation of surface evapo-ration by dust over the south-west Pacific occurs via surface radiative forcing and dust-induced stabilisation of the bound-ary layer. The results suggest that (1) a realistic treatment of Australian dust may be necessary for accurate simulation of the ENSO-rainfall relationship over Australia, and (2) radia-tive feedbacks involving dust may be important for under-standing natural rainfall variability over Australia.
Full-text available
We assess the simulation of Australian mean climate and rainfall variability in a new version of the CSIRO coupled ocean-atmosphere global climate model (GCM). The new version, called Mark 3.6 (Mk3.6), differs from its recent predecessors (Mk3.0 and Mk3.5) by inclusion of an interactive aerosol scheme, which treats sulfate, dust, sea salt and carbonaceous aerosol. Other changes include an updated radiation scheme and a modified boundary-layer treatment. Comparison of the mean summer and winter climate simulations in Mk3.6 with those in Mk3.0 and Mk3.5 shows several improvements in the new version, especially regarding winter rainfall and sea-level pressure. The improved simulation of Australian mean seasonal climate is confirmed by calculation of a non-dimensional skill score (the 'M-statistic'), using data from all four seasons. However, the most dramatic improvement occurs in the model's simulation of the leading modes of annual rainfall variability, which we assess using empirical orthogonal teleconnections (EOTs). Compared to its predecessors and several international GCMs, Mk3.6 is best able to capture the spatial pattern of the leading rainfall mode, which represents variability due to the El Nino Southern Oscillation (ENSO). Mk3.6 is also best able to capture the spatial pattern of the second rainfall mode, which corresponds to increased rainfall in the northwest, and decreased rainfall over eastern Australia. We propose a possible mechanism for the improved simulation of rainfall variability in terms of the role of interactive dust in Mk3.6. By further suppressing convection over eastern Australia during El Nino events, dust feedbacks may enhance rainfall variability there, in tune with the model's ENSO cycle. This suggests that an interactive aerosol treatment may be important in a GCM used for the study of Australian climate change and variability. Mechanistic sensitivity studies are needed to further evaluate this hypothesis.
Full-text available
EXTENDED ABSTRACT The Intergovernmental Panel on Climate Change (IPCC) has recently completed its Fourth Assessment Report (AR4) "Climate Change 2007". Results from comprehensive numerical models of the climate system are fundamentally important for understanding climate processes and how climate has changed in the past and may change in the future. Some time ago the C o m m o n w e a l t h S c i e n t i f i c a n d I n d u s t r i a l R e s e a r c h O r g a n i s a t i o n (CSIRO) completed its submission to the IPCC AR4 Model database a set of experiments simulating past, present and future climate with the Mk3 Climate System. The Mk3 model has been in development and used for production climate runs for the best part of a decade and is the end result of a significant commitment of financial and intellectual resources from a relatively small group of developers and stakeholders. The task of processing, deriving, validating and submitting Mk3 output provided significant computing and logistic challenges. Data requirements were substantially more demanding than ever before and the schedule for inclusion in the AR4 database was extremely tight. Particularly when several key model experiments were still underway when the call for data (from Joint Scientific Committee (JSC)/Climate Variability and Predictability (CVP) Working Group on Coupled Models (WGCM)) came. However, the importance of contributing to the IPCC AR4 with a climate model developed in the Southern Hemisphere cannot be underestimated. This effort will provide a useful "model development" yardstick for the Australian Community Climate Earth-System Simulator (ACCESS) development program that is underway in Australia. The best indicator of the Mk3 outcome is the inclusion of its performance in hundreds of peer-reviewed scientific articles and the many contractual reports completed and underway. We briefly describe the system for managing Mk3 model output based on a modern, locally developed, scripting computer language for the efficient processing of large and complex datasets. This system takes into account different Mk3 model configurations and model output inconsistencies and is able to generate a temporally, spatially and physically consistent set of data products. An essential feature is the ability to make the model results self describing (CF-compliant netCDF files) to enable efficient uptake by researchers. Experiences with data validation and quality control checking is described, an often overlooked aspect of data delivery. Our goal was to make the Mk3 model output easily accessible to the international climate research community. In this presentation a brief history of the development and features of the CSIRO Mk3 model will be provided. Although the Mk3.5 model was not included in AR4, the Mk3.5 version of the climate model includes many improvements over its predecessor resulting in a control climate with a relatively small drift. Mk3.5 output forms part of the new World Climate Research Programme (WCRP) Coupled Model Intercomparison Project Version 3 (CMIP3) Multimodel Dataset. This new and updated archive of model outputs will provide the local and international community with a wealth of informative and sophisticated climate model experiments for completion of important institutional studies. An indication of the regional demands for Mk3 outputs will be given. Details of how to access the CSIRO Mk3.0 and Mk3.5 data will be provided.
Full-text available
The historical surface temperature data set HadCRUT provides a record of surface temperature trends and variability since 1850. A new version of this data set, HadCRUT3, has been produced, benefiting from recent improvements to the sea surface temperature data set which forms its marine component, and from improvements to the station records which provide the land data. A comprehensive set of uncertainty estimates has been derived to accompany the data: Estimates of measurement and sampling error, temperature bias effects, and the effect of limited observational coverage on large-scale averages have all been made. Since the mid twentieth century the uncertainties in global and hemispheric mean temperatures are small, and the temperature increase greatly exceeds its uncertainty. In earlier periods the uncertainties are larger, but the temperature increase over the twentieth century is still significantly larger than its uncertainty.
Full-text available
In this study, we present climate change projections based on the results from 23 climate model simulations performed for the IPCC 4th Assessment Report. Statistical methods are used to test how well each model simulated observed average (1961-1990) patterns of mean sea-level pressure, temperature and rainfall over the Australian region. The 15 models with the highest pattern correlations and smallest rms errors are identified. The 21st century simulations are driven by the IPCC 'SRES' greenhouse gas and aerosol emission scenarios. Using the 15 best climate models, annual and seasonal average projections of Australian rainfall and temperature change are derived for various decades. Results are highlighted for 2030 and 2070 for comparison with projections published by CSIRO in 2001. The projections are expressed as ranges, incorporating uncertainty in both global warming and regional differences between climate simulations over Australia. Inland regions show greater warming, compared to coastal regions. There are large decreases in the number of days below 0°C and large increases in the number of days above 35°C or 40°C. Rainfall changes are more complex than temperature changes. Although increases and decreases in rainfall are projected in the future, decreases dominate the overall pattern, especially in the south in winter and spring. CSIRO's earlier projections, based on nine climate models, appear robust when compared with the updated projections. The patterns and magnitudes of warming are similar, although the updated projections have slightly less warming in coastal regions. The pattern of rainfall change is also similar, particularly the strong decrease in winter and spring over southern Australia, but the updated projections give a more widespread tendency for increases in summer in eastern Australia and a clearer tendency for decreases in autumn in Queensland and the eastern Northern Territory.
Full-text available
Average dust emissions from Australia are small compared to those from the major sources in the Northern Hemisphere. However, they are highly episodic, and this may increase the importance of Australian dust as a climate feedback agent. We compare two 160-year coupled atmosphere-ocean simulations of modern-day climate using the CSIRO Mark 3.6 global climate model (GCM). The first run (DUST) includes an interactive treatment of mineral dust and its direct radiative effects. The second run (NODUST) is otherwise identical, but has the Australian dust source set to zero. We focus on the austral spring season, when the correlation between rainfall and the El Niño Southern Oscillation (ENSO) is strongest over Australia. We find that the ENSO-rainfall relationship over eastern Australia is stronger in the DUST run: dry (El Niño) years tend to be drier, and wet (La Niña) years wetter. The ENSO-rainfall relationship is also weaker over north-western Australia in the DUST run. The amplification of ENSO-related rainfall variability over eastern Australia and the weaker ENSO-rainfall relationship over the north-west both represent an improvement relative to observations. The suggested mechanism over eastern Australia involves stabilisation of the surface layer due to enhanced atmospheric heating and surface cooling in El Niño years, and enhanced ascent and moisture convergence driven by atmospheric heating in La Niña years. The results suggest that (1) a realistic treatment of Australian dust may be necessary for accurate simulation of the ENSO-rainfall relationship over Australia, and (2) radiative feedbacks involving dust may be important for understanding natural rainfall variability over Australia.
In this paper, we describe a new high-quality set of historical and ongoing real- time climate analyses for Australia. These analyses have been developed for im- proving the definition of past climate variability and change over Australia and to improve on estimates of recent climate. The climate analyses cover the variables of rainfall, temperature (maximum and minimum) as well as vapour pressure at daily and monthly timescales and are complemented by remotely sensed and model- derived data described elsewhere. New robust topography-resolving analysis methods have been developed and applied to in situ observations of rainfall, temperature and vapour pressure to pro- duce analyses at a resolution of 0.05° × 0.05° (approximately 5 km × 5 km). The new methodologies are similar to those applied internationally, but in applying them to Australia we found it necessary and desirable to introduce a number of innova- tions. The resulting analyses represent substantial improvements on operational analyses currently produced by the Australian Bureau of Meteorology, and have a number of advantages over other similar data-sets currently available. Careful attention has been paid to developing systems and data-sets which are robust and useful for the monitoring of both climate variability and climate change. These systems are now running in real time and are expected to form the basis for the ongoing monitoring of Australia's surface climate variability and change by the Australian Bureau of Meteorology. The underlying data and associated error sur- faces (grids and station data) are updated in real time and are all available free of charge through the Bureau's climate website (
An overview is given of the Australian Commonwealth