To many, a storm warrants no more thought than the effort it takes to open an umbrella.
But for the people working round the clock at Singapore's national weather office, no two downpours are the same.
"To the public, rain is just rain," said Ms Patricia Ee, 52, director of weather services at the Meteorological Service Singapore (MSS).
"But when a meteorologist sees rain today, tomorrow or the day after, he sees them as different weather systems.
"In that sense, it is never boring."
Strong upper-level winds, for example, can prevent rain clouds from rising high and, as a result, the country experiences showers rather than thundery showers.
A meteorologist would want to understand why the weather is behaving that way and how it is interacting with the environment, she said.
"Supposing we expect a thunderstorm in our forecast to the aviation industry, we will tell it the expected visibility, what the wind conditions will be and whether there will be gusts. These are essential for flights to come in safely and efficiently."
Ms Ee leads a department with a staff of more than 90 at the National Environment Agency which is responsible for forecasting Singapore's weather for the public, the aviation industry and even the military.
The team also monitors seismic activities or volcanic eruptions where the emission of volcanic ash could affect flights coming into and leaving Singapore.
Coming up with a weather forecast is a complicated process which requires a whole suite of monitoring tools such as weather radars and satellites, wind profilers and surface sensors, said Ms Ee, who has been in the field for over 20 years. Data from the upper layers of the atmosphere is also collected using a weather balloon.
But collecting the data is just the first step. The team still needs to analyse all the information and make a decision on what to expect next.
-
75%
Percentage of accurate weather predictions for 24-hour forecasts.
90%
Percentage of accurate weather predictions for two-hour forecasts.
"You have the public and other agencies calling to ask for the weather forecast. When the call comes in, you cannot say 'just hold on, I'll go and check'," she said.
Advances in technology have made it easier for the team.
-
Ask the expert
-
The Meteorological Service Singapore at the National Environment Agency answers two common questions about the weather:
Q Is it raining more in Singapore?
A From 1980 to last year, the annual rainfall over Singapore has risen on average by 10.1mm per year.
This is a small amount when set against the long-term mean annual rainfall of around 2,170mm.
There are major uncertainties in attributing observed rainfall trends to particular factors, either natural or man-made. It is also important to note that recent rainfall trends are not an indicator of future trends.
Over the long term, while wetter and drier periods are projected to get wetter and drier respectively, our annual rainfall is not expected to change significantly.
However, consistent with a warmer future climate, the intensity and frequency of heavy rainfall events are expected to increase.
Q How accurate is our local weather forecast?
A On average, the accuracy of our weather forecasts ranges between 75 per cent for 24-hour forecasts and 90 per cent for near-term two-hour nowcasts.
While there have certainly been occasions when our weather forecasts were not accurate, there have been many more occasions when the forecasts were "spot on" but perceived by some people as being inaccurate.
This is more an issue of interpretation and expectation. For example, when showers fall over Bedok and Marine Parade and not Tampines, someone in Tampines would likely interpret a forecast of "thundery showers over the eastern part of Singapore" as inaccurate.
Nonetheless, we are exploring ways to better communicate our forecasts to users to minimise misinterpretation.
A perfect forecast is not possible for a number of reasons, particularly because the state of the atmosphere at the start of the forecast cannot be known with absolute accuracy. Tiny errors in the initial state can be amplified such that, after a while, the forecast becomes quite useless.
Typically, the main weather systems in mid-latitudes, which tend to be large scale, can be forecast reasonably well up to a week in advance. However, in the tropics where transient, small-scale thunderstorms are dominant, forecasts from numerical weather-prediction models are reliable for just one to three days ahead.
For instance, Japan's Himawari-8 geostationary satellite, launched in 2014, has brought significant improvements to how the team monitors the weather, including the development of thunderstorms, movement of transboundary smoke haze and, at times, volcanic ash.
It gives the team access to satellite images every 10 minutes, with improved resolution as high as 0.5km, compared with hourly observations with a resolution of 4km that it used to receive.
Nonetheless, the weather predictions are not always spot on. They are accurate 75 per cent of the time for 24-hour forecasts and 90 per cent for two-hour nowcasts, or very short-term forecasts.
To this end, the MSS' Centre for Climate Research is working with the British national weather service to develop the world's first numerical weather-prediction model for the tropics - a project that started in 2013. The model is a software that uses weather data, such as rainfall and wind speeds, to make a forecast.
Current models are more suited to the higher latitudes, where the weather systems are of a larger scale and can be monitored and tracked for days. In Singapore, most systems are small and can develop quickly overhead before disappearing just as quickly. That is what makes the weather here difficult to forecast, explained Ms Ee. She added that such models can still have significant errors due to the chaotic nature of the atmosphere.
She said: "Very small errors in the initial conditions can develop into large errors in the forecast. This is the so-called 'butterfly effect'.
"It is very satisfying every time we get the forecast right."
When World Meteorological Organisation secretary-general Petteri Taalas was here last month, he explained why predicting weather in this region is challenging.
He said that as weather patterns span international boundaries, and as one needs expensive ground-and ocean-based observation systems across huge distances to provide accurate data, poorer countries that cannot afford this technology would have only sparse observation networks. This would lead to missing pieces of information and therefore less accurate predictions.
Climate change would only make predicting droughts, floods and the seasons harder, he said. "Weather is already becoming more unpredictable and extreme because of climate change. We expect this trend to continue for the next 50 years."
Ms Ee and her team also have to handle 10 to 20 phone calls a day from the public. Some calls can be rather unusual, such as one she received about 10 years ago.
"The caller told me, 'I need to know if the weather will be sunny tomorrow because I need to dry all my vegetables to make achar for Chinese New Year,'" said Ms Ee.
"We will try to answer their questions... People do call to complain as well, which motivates us to help them to understand the challenges we face... and to improve our weather forecasts," she said.
- Additional reporting by Jose Hong
