This is one in a 2014 guest column series which builds on the one in 2009 where 50+ had written about how science/tech has evolved their hobby/interest.
This time it is Simon Griffiths. Based in Johannesburg, South Africa he has been in the enterprise software industry for 20 years. He works on global product marketing for international ERP vendor, SYSPRO. And on his gauge at home where he has recorded daily rainfall just almost as long.
To most people the word "meteorology" conjures up images of a TV presenter talking about current weather conditions and the forecast for the next few days. That portrays the wrong image of meteorology, it’s meteorology as entertainment rather than the science that it is.
Human societies have tried to forecast the weather probably since they started cultivating crops. The subject of meteorology was begun as far back as 350 BC when Aristotle gave it its name. The history of meteorology is an intriguing subject and .">.">summarized well. There are different branches of the subject – e.g., maritime, agricultural.
My introduction to meteorology was as a subject a university, but it was never something I considered too interesting. My views began to change in my final year when one of my lecturers suggested that there was a research job for a company that was seeding clouds in eastern South Africa. As I had no other employment lined up, I applied for the job, and shortly afterwards, got it. That was in 1979.
The job brought me closer to meteorology than ever before – it was a case of academic discussions becoming practical. I launched weather balloons, tracked and recorded the soundings from the transmitter on the balloons, and got to understand the process of forecasting from the company meteorologist. During the summer rainy season I worked on the weather radar and watched how clouds moved and developed in real-time. I even found it interesting to learn about the intricacies of cloud physics, which had been too theoretical for me before. Within 18 months, I could act like an expert and gauge the distance of a storm just by looking at it.
While I was there, and afterwards, the company was featured in several media reports about its cloud seeding work, unfortunately the only website that still has a valid link is to a BBC TV program called The Rainmaker. The Economist also wrote an article called "Cloudbusting", published on 21st August 1999. There were later academic articles about it, such as a report in the American Meteorological Society journal in September 1996.
It was also where I got my first exposure to computing. The company was very data-oriented – as much as it could be then. Its weather radar was linked to a Data General mini-computer where we could record and store the profile of thunderstorms. A particle-measuring laser probe was added to the wingtip of one of the aircraft to get detailed data on atmospheric particles. From one of the particle scans we could clearly see a snow crystal, which lead me to joke that I had seen snow in summer. I learnt to program there, writing programs in Basic and FORTRAN, as a kind of apprentice to the technical director, an ex-US Navy and NASA engineer. During my time the computer was upgraded, and we got new removable disk drives that could hold a whopping five megabytes of data – a lot in those days.
While I was working there in 1980, I was told about one of the great technological marvels of meteorology, the Cray super-computer at NCAR (National Center for Atmospheric Research) in Boulder, Colorado. This was a machine that cost millions of dollars and could compute calculations at a then incredible rate; and it was housed in a special super-cooled environment. These days, the average smartphone has more computing power than that Cray.
The fact that I was introduced to computers while doing meteorological work should not come as a great surprise. Meteorology and technology have been closely linked in recent decades. Computing power is what it takes to do the mass of complex calculations required for a forecast. The two main elements of forecasting are data from observed measurements (wind, temperature, pressure), and the application of physical laws of the atmosphere. A further complication is that while the atmosphere is governed by the laws of physics, the Earth’s weather is a chaotic, non-linear system – in simple English that means it is highly sensitive to initial conditions, and the output is not proportional to the input. That is why a major challenge has always been making accurate weather forecasts for a reasonable period into the future.
Until the advent of computers, it was practically impossible to do this to any great scale. These days we take it for granted that major weather forecasting services use computers to do their numerical forecasting.
Another revolution has been the use of satellite imagery to back up weather forecasting. Until the NOAA (National Oceanic and Atmospheric Administration) satellites of the 1970s, we could only observe the atmosphere via land- or sea-based weather stations, and then create synoptic charts as a guide to what the weather was doing. Now we can use both to show what the weather is like. This was recently born out to me when there was a warning of a major cold front approaching South Africa.
Only a few years ago, we got our weather news from the radio, TV or newspapers. Now people have apps on their smartphones like AccuWeather and can get hourly information geo-located to their position.
Despite all the incredible technological advances in the last fifty years, meteorologists can still only forecast with any accuracy for a few days into the future, and are known to get even that wrong. I wonder what advances will happen in the next fifty years.