Why we need old weather data

When people ask me what I do here, my standard response is “Soy investigadora, en el Centro de Cambio Climatic”. Most people take this to mean that I work with the political and economic solutions required to solve the diabolical problem of climate change (which they then quiz me about), but sadly this is not true.

I work with old weather.

Yep, old numbers. Historical weather observations taken up to 250 years ago. I find them, digitise them, and check them to see how reliable they are.

When I occasionally manage to explain this in my basic Spanish, people generally look disappointed, confused, and then they slink away.

Recovering old weather data is not at the “coal face” of climate change research (haha, pun), and many people may think that it’s not really important for helping us figure out how we are going to manage the future.

How wrong they are!

Here are four reasons why historical weather observations are so useful, and why, really, we need them more than ever.


Without historical weather observations, our understanding of extreme events — the aspect of climate change that has people most frightened — is obscured. How do we know if the Australian drought in the 1990s and 2000s is more or less severe than the one at the turn of the 20th century? How can we prepare cities and buildings for a storm without understanding what previous storms have done?

Old weather observations help place recent extreme events into a long-term context, giving us detailed information about weather patterns on a daily or monthly resolution. You can’t get this kind of accuracy from palaeoclimate records, and you can’t be overly trusting of written records either in a long-term context: people’s memories can be notoriously short. There’s always one guy who says that a storm is “the worst he’s ever seen”, even if that is not strictly true. So the numbers really matter.


Without good coverage of weather observations in space, and time, our hands are tied when it comes to a) quantifying climate change that has happened, and b) modelling what might happen further down the track.

In the Southern Hemisphere, temperature estimates from University of East Anglia’s Climate Research Unit are based on 100 stations before about 1900. Only 100 stations, for the whole hemisphere! While very clever algorithms are used to extrapolate the data for global and hemispheric temperature values, that is not a lot of observations to represent the bottom half of the earth.

The same thing can be said for reanalysis products, which use observations and physics to create detailed models of how the atmosphere behaves. The historical 20th Century Reanalysis product for example goes back to 1850 uses sub-daily atmospheric pressure observations. Again, the uncertainty in the Southern Hemisphere is quite large before the mid-20th century, because of a lack of data.

Lots of weather observations that cover a wide area and go a long way back in time are the only things that can improve both of these issues. Any data that can be found therefore, especially for poorly represented areas such Africa, Antarctica or the ocean, can greatly improve the accuracy of global and regional climate products.


Long records of temperature and rainfall are uber-important for palaeoclimatology studies too. Scientists who extract the climate signal from tree rings, ice cores, coral skeletons and cave deposits need long, reliable instrumental climate records to calibrate their natural records, and match the squiggles they see in their data to real climate variations. The longer the climate records are, the better the calibration will be.

Recently, for example, Lorrey and Chappell used a 19th century weather diary in conjunction with New Zealand tree-ring records to show that it was colder than normal in the north of New Zealand during the early to mid 19th century. It was the power of these two sources combined that allowed them to determine their results with increased certainty. We’ve also used long-term rainfall records to assess the reliability of a palaeoclimate rainfall reconstruction for southeastern Australia.


Written accounts of the weather and climate, and how past societies adapted to changes, are becoming increasingly important sources as we grapple with how to handle living with a changing climate.

The relationship between how the weather affected and was experienced by societies, and what happened in a meteorological sense, can provide important verification for both the qualitative and quantitative data. In other words, historical numbers and words can really help to improve understanding of each other.

A recent article about the international data rescue initiative Atmospheric Climate Reconstructions over the Earth (ACRE) gives some fascinating examples of how historical weather observations and documentary accounts have been used together to explore things like extreme events in the UK and snowfall in Wales.

Another example is missionary diaries from Lesotho in the 19th century, which were used to identify temperature variations in the region, with the help of some recovered instrumental data. There are more calls now for historians, physical scientists, and social scientists to use historical data to learn how humans have adapted and responded to previous changes in their environment.


So you see? They may sound old and dusty, but historical weather records really have a lot to tell us about the past, the present, and the future.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: