Logo
Print

As has been previously stated, climate change scenarios are developed using climate models (UNFCCC). These models use mathematical representations of the climate system, simulating the physical and dynamical processes that determine global/regional climate. They range from simple, one-dimensional models to more complex ones such as global climate models (known as GCMs), which model the atmosphere and oceans, and their interactions with land surfaces. They also model change on a regional scale (referred to as regional climate models), typically estimating change in areas in grid boxes that are approximately several hundred kilometers wide. It should be noted that GCMs/RCMs provide only an average change in climate for each grid box, although realistically climates can vary considerably within each grid. Climate models used to develop climate change scenarios are run using different forcings such as the changing greenhouse gas concentrations. These emission scenarios known as the SRES (Special Report on Emission Scenarios) developed by the Intergovernmental Panel on Climate Change (IPCC) to give the range of plausible future climate. These emission scenarios cover a range of demographic, societal, economic and technological storylines . They are also sometimes referred to as emission pathways. Table 1 presents the four different storylines (A1, A2, B1 and B2) as defined in the IPCC SRES.


Climate change is driven by factors such as changes in the atmospheric concentration of greenhouse gases and aerosols, land cover and radiation, and their combinations, which then result in what is called radiative forcing (positive or warming and negative or cooling effect). We do not know how these different drivers will specifically affect the future climate, but the model simulation will provide estimates of its plausible ranges.

A number of climate models have been used in developing climate scenarios. The capacity to do climate modeling usually resides in advanced meteorological agencies and in international research laboratories for climate modeling such as the Hadley Centre for Climate Prediction and Research of the UK Met Office (in the United kingdom), the National Center for Atmospheric Research and the Geophysical Fluid Dynamics Laboratory (in the United States), the Max Planck Institute for Meteorology (in Germany), the Canadian Centre for Climate Modeling and Analysis (in Canada), the Commonwealth Scientific and Industrial Research Organization (in Australia), the Meteorological Research Institute of the Japan Meteorological Agency (in Japan), and numerous others. These centers have been developing their climate models and continuously generate new versions of these models in order address the limitations and uncertainties inherent in models.

For the climate change scenarios in the Philippines presented in this Report, the PRECIS (Providing Regional Climates for Impact Studies) model was used. It is a PC-based regional climate model developed at the UK Met Office Hadley Centre for Climate Prediction and Research to facilitate impact, vulnerability and adaptation assessments in developing countries where capacities to do modeling are limited. Two time slices centered on 2020 (2006-2035) and 2050 (2036-2065) were used in the climate simulations using three emission scenarios; namely, the A2 (high-range emission scenario), the A1B (medium- range emission scenario) and the B2 (low-range emission scenario).

The high-range emission scenario connotes that society is based on self-reliance, with continuously growing population, a regionally-oriented economic development but with fragmented per capita economic growth and technological change. On the other hand, the mid-range emission scenario indicates a future world of very rapid economic growth, with the global population peaking in mid-century and declining thereafter and there is rapid introduction of new and more efficient technologies with energy generation balanced across all sources. The low-range emission scenario, in contrast, indicates a world with local solutions to economic, social, and environmental sustainability, with continuously increasing global population, but at a rate lower than of the high-range, intermediate levels of economic development, less rapid and more diverse technological change but oriented towards environment protection and social equity. 


To start the climate simulations or model runs, outputs (climate information) from the relatively coarse resolution GCMs are used to provide high resolution (using finer grid boxes, normally 10km-100km) climate details, through the use of downscaling techniques. Downscaling is a method that derives local to regional scale (10km-100km x 10km-100km grids) information from larger-scale models (150km-300km x 150km-300km grids) as shown in Fig.1. The smaller the grid, the finer is the resolution giving more detailed climate information.
 


The climate simulations presented in this report used boundary data that were from the ECHAM4 and HadCM3Q0 (the regional climate models used in the PRECIS model software). 

How were the downscaling techniques applied using the PRECIS model?

To run regional climate models, boundary conditions are needed in order to produce local climate scenarios. These boundary conditions are outputs of the GCMs. For the PRECIS model, the following boundary data and control runs were used:


For the high-range scenario, the GCM boundary data used was from ECHAM4. This is the 4th generation coupled ocean-atmosphere general circulation model, which uses a comprehensive parameterization package developed at the Max Planck Institute for Meteorology in Hamburg, Germany. Downscaling was to a grid resolution of 25km x 25km; thus, allowing more detailed regional information of the projected climate. Simulated baseline climate used for evaluation of the models capacity of reproducing present climate was the 1971-2000 model run. Its outputs were compared with the 1971-2000 observed values.
For the mid-range scenario, the GCM boundary data was from the HadCM3Q0 version 3 of the coupled model developed at the Hadley Centre. Downscaling was also to a grid resolution of 25km x 25km and the same validation process was undertaken.
For running the low-range scenario, the same ECHAM4 model was used. However, the validation process was only for the period of 1989 to 2000 because the available GCM boundary data in the model was limited to this period.
The simulations for all 3 scenarios were for three periods; 1971 to 2000, 2020 and 2050. The period 1971 to 2000 simulation is referred to as the baseline climate, outputs of which are used to evaluate the models capacity of reproducing present climate (in other words, the control run). By comparing the outputs (i.e., temperature and rainfall) with the observed values for the 1971 to 2000 period, the models ability to realistically represent the regional climatological features within the country is verified. The differences between the outputs and the observed values are called the biases of the model. The 2020 and 2050 outputs are then mathematically corrected, based on the comparison of the models performance.

The main outputs of the simulations for the three SRES scenarios (high-range, mid-range and low-range) are the following:
projected changes in seasonal and annual mean temperature;
projected changes in minimum and maximum temperatures;
projected changes in seasonal rainfall; and
projected frequency of extreme events.
The seasonal variations are as follows:
the DJF (December, January, February or northeast monsoon locally known as amihan) season;
the MAM (March, April, May or summer) season;
the JJA (June, July, August or southwest monsoon season, or habagat) season; and
the SON (September, October, November or transition from southwest to northeast monsoon) season.


On the other hand, extreme events are defined as follows:
extreme temperature (assessed as number of days with maximum temperature greater than 35 ยบC, following the threshold values used in other countries in the Asia Pacific region);
dry days (assessed as number of dry days or day with rainfall equal or less than 2.5mm/day, following the World Meteorological Organization standard definition of dry days used in a number of countries); and
extreme rainfall (assessed as number of days with daily rainfall greater than 300mm, which for wet tropical areas, like the Philippines, is considerably intense that could trigger disastrous events).

How were the uncertainties in the modeling simulations dealt with?

Modeling of our future climate always entails uncertainties. These are inherent in each step in the simulations/modeling done because of a number of reasons. Firstly, emissions scenarios are uncertain. Predicting emissions is largely dependent on how we can predict human behavior, such as changes in population, economic growth, technology, energy availability and national and international policies (which include predicting results of the international negotiations on reducing greenhouse gas emissions). Secondly, current understanding of the carbon cycle and of sources and sinks of non-carbon greenhouse gases are still incomplete. Thirdly, consideration of very complex feedback processes in the climate system in the climate models used can also contribute to the uncertainties in the outputs generated as these could not be adequately represented in the models.


But while it is difficult to predict global greenhouse gas emission rates far into the future, it is stressed that projections for up to 2050 show little variation between different emission scenarios, as these near-term changes in climate are strongly affected by greenhouse gases that have already been emitted and will stay in the atmosphere for the next 50 years. Hence, for projections for the near-term until 2065, outputs of the mid-range emission scenario are presented in detail in this Report.

Ideally, numerous climate models and a number of the emission scenarios provided in the SRES should be used in developing the climate change scenarios in order to account for the limitations in each of the models used, and the numerous ways global greenhouse gas emissions would go. The different model outputs should then be analyzed to calculate the median of the future climate projections in the selected time slices. By running more climate models for each emission scenarios, the higher is the statistical confidence in the resulting projections as these constitute the ensemble representing the median values of the model outputs.

The climate projections for the three emission scenarios were obtained using the PRECIS model only due to several constraints and limitations. These constraints and limitations are:

Access to climate models: at the start, PAGASA had not accessed climate models due to computing and technical capacity requirements needed to run them;
Time constraints: the use of currently available computers required substantial computing time to run the models (measured in weeks and months). This had been partly addressed under the capacity upgrading initiatives being implemented by the MDGF Joint Programme which include procurement of more powerful computers and acquiring new downscaling techniques. Improved equipment and new techniques have reduced the computing time requirements to run the models. However, additional time is still needed to run the models using newly acquired downscaling techniques; and

The PAGASA strives to improve confidence in the climate projections and is continuously exerting efforts to upgrade its technical capacities and capabilities. Models are run as soon as these are acquired with the end-goal of producing an ensemble of the projections. Updates on the projections, including comparisons with the current results, will be provided as soon as these are available.  

What is the level of confidence in the climate projections?

The IPCC stresses that there is a large degree of uncertainty in predicting what the future world will be despite taking into account all reasonable future developments. Nevertheless, there is high confidence in the occurrence of global warming due to emissions of greenhouse gases caused by humans, as affirmed in the IPCC Fourth Assessment Report (AR4). Global climate simulations done to project climate scenarios until the end of the 21st century indicate that, although there are vast differences between the various scenarios, the values of temperature increase begin to diverge only after the middle of this century (shown in Fig.3). The long lifetimes of the greenhouse gases (in particular, that of carbon dioxide) already in the atmosphere is the reason for this behavior of this climate response to largely varying emission scenarios. 

Model outputs that represent the plausible local climate scenarios in this Report are indicative to the extent that they reflect the large-scale changes (in the regional climate model used) modified by the projected local conditions in the country.

It also should be stressed further that confidence in the climate change information depends on the variable being considered (e.g., temperature increase, rainfall change, extreme event indices, etc.). In all the model runs regardless of emission scenarios used, there is greater confidence in the projections of mean temperature than that of the others. On the other hand, projections of rainfall and extreme events entail consideration of convective processes which are inherently complex, and thus, limiting the degree of confidence in the outputs. 

What are the possible applications of these model-generated climate scenarios?

Climate scenarios are commonly required in climate change impact, vulnerability and adaptation assessments to provide alternative views of future conditions considered likely to affect society, systems and sectors, including a quantification of climate risks, challenges and opportunities. Climate scenario outputs could be used in any of the following:

to illustrate projected climate change in a given administrative region/province;
to provide data for impact/adaptation assessment studies;
to communicate potential consequences of climate change (e.g., specifying a future changed climate to estimate potential shifts in say, vegetation, species threatened or at risk of extinction, etc.); and
for strategic planning (e.g., quantifying projected sea level rise and other climate changes for the design of coastal infrastructure/defenses such as sea walls, etc.).