Inaccurate weather data causes problems

Automation is a great way to reduce costs, but without human oversight it can create problems.

An example is an Environment Canada automated weather station in Regina that is giving the impression that precipitation this winter is only 72 percent of the long-term average.

That is a head scratcher to Regina-area farmers who have plowed their yards for the umpteenth time this winter.

Actually, the area has record high snowfall, a concern come spring.

This is not a flaw limited to the Regina station. This defect has been brought to light before.

A report calledDegradation in Environment Canada’s Climate Network, Quality Control and Data Storage Practices: A Call to Repair the Damageilluminated this and other troubles.

The report says budget cutting in the 1990s at the department’s meteorological service and continued underfunding led to a rapid change from human observers to automated sensors. The sensors work well for factors such as temperature and humidity, but have problems with measurements such as precipitation accumulation, snow depth and visibility.

If snow falls on a calm day the machine catches it, but if there is a wind blowing it can miss it.

So the country-wide network of Environment Canada’s automated weather stations gets a failing grade when it comes to measuring snow.

Erroneous readings from automated stations do more than hurt the credibility of the weather service, the report said.

The data goes into the climate archive and if errors are not corrected, it can lead to misunderstandings, erroneous climate research and damaging public policy.

For example, the inaccurate snowfall data from Regina appeared in Agriculture Canada’s drought monitoring maps.

Farmers know how much snow covers their fields, but government policy planners and market analysts in far away cities using the big picture from these maps could have an inaccurate impression.

Given the shortfalls of the automated systems, it might be advisable to simply not publish the snow data.

However, there is a great need and desire to get accurate data.

The Prairies face important water challenges. Growing urban populations, industrial growth, livestock production and irrigation are creating water demands that strain some river systems in this semi arid region.

Most climate scientists agree the climate is changing. For the Canadian Prairies, and the Rocky Mountains where our rivers originate, computer models predict the amount and timing of precipitation will change.

These predictions must be assessed with accurate long-term monitoring to help policy makers and citizens react and plan appropriately.

Luckily, Environment Canada does not rely exclusively on automated weather station data. In some places government employees and volunteers collect data and send it to Environment Canada.

But their numbers are dwindling. AToronto Starstory said in the late 1980s there were about 2,500 volunteer observation stations. Now there are about 630, indicating a need for a revival of this valuable volunteer resource.

Satellites also provide important monitoring. However, Canadians have a growing belief that Environment Canada’s weather service is falling behind.

Many turn to private forecasters and, in agriculture, farmers are buying weather stations and a growing number are willing to pay to get useful forecasts.

But private forecasters cannot replace the need for a national weather service with the tools, funding and staff needed to collect and archive accurate, publicly accessible data needed to understand and prosper in our climate.

Decisions are only as good as the information they are based on.

Bruce Dyck, Terry Fries, Barb Glen and D’Arce McMillan collaborate in the writing of Western Producer editorials.


Stories from our other publications