Uncertainties and Scientific and
Technological Challenges in Weather and Climate Forecasting
Shiromani Jayawardena
Deputy Director (Research and Climate Change), Department
of Meteorology
Over the last few decades, the
frequency and intensity of extreme climate events have been increasing,
especially due to global warming caused by human activities. According to a special
report by the Intergovernmental Panel on Climate Change 2012, changes in the
frequency, intensity, spatial extent, duration, and timing of weather and
climate extremes can increase people’s vulnerability to natural disasters and
become a major threat to global economies. Weather forecasts, ranging from timescales
of few hours to entire seasons, can reduce susceptibility to weather variations
and climate-related disasters, improve food security and health outcomes, and
enhance water resource management. However, a major challenge is predicting future
weather patterns in a rapidly changing climate. The future is intrinsically
uncertain. Further, dynamic physical processes of atmosphere and its
interactions with surrounding systems (e.g. land, ocean, and ice surfaces) make
forecasting even more difficult.
Forecasting Future Weather Patterns
Weather forecasts are prepared in
a systematic way, which involves observation, process understanding,
prediction, and dissemination. Each of these components has benefitted, and
will continue to benefit from advances in science and technology. Over the past few decades, substantial leaps
in science have resulted in improved and more efficient methods of making
timely observations, using a wide variety of sources including radars and
satellites. The use of such observation tools has dramatically increased the
quality, value, and reliability of weather forecasts, assisting decision-making
processes around the world. The impact of improved observational capability,
increased scientific understanding of atmospheric processes, more sophisticated
computer resources, skillful numerical models, and other forecasting tools have
gradually changed the public’s perception of the accuracy of weather forecasts.
Two areas that achieved significant improvements in this regard are numerical
weather forecasting and seasonal predictions.
Numerical Weather Prediction (NWP)
Numerical Weather Prediction
(NWP) is a forecasting method that uses complex mathematical equations, based
on well-established physical laws, to predict atmospheric behaviour. This is
done by undertaking complex mathematical operations that model the current
state of atmosphere, obtained through observations. High speed, high-performing
computers are required to carry out the enormous number of calculations
involved in this process. To get the current
status of the atmosphere, routine and accurate measurements at ground and upper
air stations, as well as remote sensing systems, are required.
Seasonal Predictions
Seasonal predictions are made in
probabilistic terms, such as probability of receiving average/above
average/below average level of rainfall/temperature over a season. Currently,
seasonal predictions are made using both statistical schemes and dynamical
models. The statistical approach seeks to find recurring patterns in climate,
associated with a ‘predictor field’ such as sea surface temperature. Such
models have demonstrated the ability to forecast El Niño and some of its global
climate impacts. The basic tools for dynamic prediction include both
atmospheric and oceanic models.
Challenges in Weather Forecasting
Whatever the successes, some
level of uncertainty remains and there are challenges that act as barriers
against further progress in weather forecasting.
Nonlinearity of atmospheric systems: Successful weather forecasts are possible if the processes are
understood properly, and if the current state of the atmosphere is well-known. Even
though, the scientific understanding of atmospheric and oceanic systems has
made considerable progress through a variety of research activities, including
field experiments, theoretical work, and numerical simulations, atmospheric
processes are inherently non-linear. Furthermore, all physical processes cannot
be understood or represented in NWP models. Continued research efforts, using
computer technology and physical measurements, will improve these
approximations. Even then, it will not be possible to represent all atmospheric
motions and processes in NWPs, especially over tropical regions.
Limitations in observations: Despite recent advances, there are
limitations in making observations, especially in desert areas and oceans. As a result, there is a need for improved
observation systems and methods to assimilate these data into NWP models.
Predictability desert - dub-seasonal prediction:
Substantial progress has been made in recent years in the development and
application of short-range to medium-range weather forecasts and seasonal
climate predictions. However, forecasting on the time scale beyond 14 days to
season (sub-seasonal time scale) that lies between daily weather and seasonal
climate has received much less attention because it is considered a
‘predictability desert’. Sub-seasonal to seasonal prediction is a crucial
planning window for the agricultural sector, water resource management, and
other stakeholders, such as transport planners. There is a growing interest in
the scientific community to develop forecasts that would fill the gap between
medium-range weather forecasts and long-range or seasonal ones. This interest
in sub-seasonal prediction was triggered, not only by a growing demand from potential
users, but also from the progress in areas such as medium-range forecasting, as
well as the predictive sources like Madden Julian Oscillation (MJO), over the
past decades
Weather Forecasting in Sri Lanka
Sri Lanka has major challenges in
forecasting weather, due to its geographical location. Sri Lanka is close to
the equator and is surrounded by the Indian Ocean, where observation density is
sparse. Some atmospheric processes in the tropical region are not fully
understood and impossible to resolve due to technology constraints. Data
sparseness in the surrounding ocean, lack of scientific understanding of some
weather systems inherent to the tropical region, and frequent, vigorous changes
in the equatorial tropical atmosphere hinder the accuracy of weather
forecasting in Sri Lanka. It is important to understand that, even with
advancements in science, some meteorological phenomena associated with extreme
weather events will remain inherently unpredictable.
However, the advances in
computational technology, together with improvements in physical representation
of the main atmospheric processes, have increased the accuracy of forecasts. It
has prompted the use of short-range NWP forecasts and seasonal forecasts in
many activities, dependent on weather and climate.
The Department of Meteorology
(DoM) runs the Weather Research and Forecasting (WRF) model for operational
purposes, as a tool in short range and medium range forecasting. The present
system has 15x15 km2 horizontal resolutions for the outer domain, and 5x5 km2
for the inner domain covering Sri Lanka. The DoM’s goal is to provide detailed,
accurate, and reliable information for weather-dependent customers and
stakeholders. Therefore, the use of a very high-resolution limited area NWP
model, with horizontal grid size of nearly 1 km, is well justified for
countries like Sri Lanka, that have highly varying terrain from central
highlands (containing many complex topographical features such as ridges,
peaks, plateaus, basins, and valleys) to low laying plains. To achieve this
goal, institutional strengthening, capacity building, strengthening and
enhancement of the observation network, enhancement of research activities is
important. In addition, improving computer resources, such as high-end computer
servers, is needed.
The DoM currently provides
seasonal as well as monthly probabilistic rainfall and temperature forecasts, at
district levels, using statistical downscaling of global models. Monthly and seasonal climate
predictions are valuable in agricultural decision making, water resource management, and other
climate sensitive activities. It is important to provide guidance
on how to use probabilistic predictions to support these decisions. T Some
scientific research has identified that, provided with sufficient information
about level of uncertainty, people would take better risk-based decisions.
Understanding the limitations of weather and climate forecasts will result in
the improved, rational use of forecasts and other weather information by
decision makers.
No comments:
Post a Comment