Gro Rolls Out The GCI - Weather Variability

Talk to our our team about Gro's offering
Talk to our team

Climate change is introducing new levels of unpredictability to doing business worldwide because of the uncertainty of how different regions are, and will continue to be, impacted. Companies large and small are now confronted with managing this unavoidable and hard to quantify risk in their day-to-day operations and throughout their supply chains. 

To help companies manage climate uncertainty, we at Gro, set out to measure how the predictability of temperature and precipitation compares with the past trends globally.  

We found that in regions where weather has become less predictable, the type of planning and hedging strategies required to minimize disruptions are markedly different than in regions where the predictability has remained relatively stable. 

We also wanted to finetune how natural disasters factor into variation estimates, while making these estimates more responsive to recent events. 

For example, after Egypt's “dragon” storm in March 2020, the prior way of computing Gro’s Climate Indicator (GCI) - Weather Variability, formerly known as the Climate Variation Index (CVI), involved treating all the measurements in the volatility estimation window the same. This meant that our indicator perpetuated the effect of the storm and elevated volatility levels one full year after the storm before dropping the volatility level off suddenly as the storm disappeared from our rear-view mirror. To avoid this sudden volatility drop, we changed the indicator by weighting recent events more heavily in the volatility estimation; this allows for the influence of a single event to diminish as that recedes into the past. 

Through this adjustment, GCI - Weather Variability users can better identify regions across the world that are experiencing more extreme weather due to climate change. And with the updated application, users can design long-term sustainability strategies that more accurately account for potential climate impacts. As with all of Gro's climate indices, the GCI - Weather Variability includes historical data and updates daily.

Methodology behind the GCI - Weather Variability:

Our newly improved indicator, GCI - Weather Variability, measures variations in land surface temperature (LST) and precipitation variables. To capture real-time data and the impacts of climate volatility on yield production, after seasonal variability is removed, we added two new metrics:

Variation of Temperature

Variation of Precipitation 

If you would like to keep using Gro’s older indicator, which removes seasonal variation, it is still available. It’s metrics include: 

Variation of Temperature Difference from 10-yr Mean V1 

Variation of Precipitation Difference from 10-yr Mean V1

While there are some commonalities, the methodology of the Variation pair of metrics used in the new GCI - Weather Variability differs from the Variation Difference from the 10-year Mean V1 pair of metrics used in our older version in a few key aspects. 

With regards to the common aspects, both pairs of metrics seek to capture day-to-day variations and both remove the variation linked to seasonal cycles. Also, both use a one-year lookback window to estimate variability. This means that events that happened more than a year before a given point in time will not influence the variability estimate for that point in time. 

Our updated GCI - Weather Variability’s pair of variation metrics, include some key differences: 

They weigh recent events more heavily when determining variability. For example, if a big storm causes an abnormal amount of rainfall that value of Variation of Temperature will be bigger than that of Variation of Temperature Difference from 10-year Mean V1 because it is more influenced by this recent abnormal jump in precipitation. 

They allow the importance of an abnormal event to decay more quickly as the event recedes in time. This means that by the time the event recedes beyond the lookback window, its importance is already negligible. (By contrast, in the older pair of metrics, the importance of the event stays the same until it drops out of the lookback window.) 

They attempt to equalize the amount of day-to-day variability that occurs in the winter versus summer. (By contrast, in the Variation Difference from 10-year Mean V1 pair of metrics, this additional step of compensating for seasonal differences in variability itself - instead of just removing the seasonal swings from the precipitation or temperature values, which both metrics do - is not taken.) By adding this step, our new GCI - Weather Variability amplifies the importance of events that are abnormal for the time of the year in which they occur. A big rainstorm occurring in the dry season is one example. 

What does this mean practically?

Let us consider a recent abnormal weather event like the catastrophic floods in Western Europe that occurred between July 12-15, 2021. One of the hardest hit areas was western Germany, particularly near the Rhine River basin. The results for Germany on July 15 can be read here

Using the updated Variation of Precipitation metrics, we can immediately see that Rhine River basin regions had significantly elevated precipitation variability. This happened because the recent abnormal event was given a high weight in the variability metric. 

By contrast, the Variation of Precipitation Difference from 10-year Mean V1 does show slightly elevated variability in western Germany, but not enough to cause a deeper shade on the color scale. This is because the influence of the recent western-tilted abnormal event on the variability metric is diluted as other events that happened in the course of the year are factored in. 

Neither metric is wrong. But the new metric highlights the most recent variability situation more, while the old metric chooses to highlight year-long differences in variation. Using both metrics together tells the full story more completely: in Germany, in mid-July 2021, year-to-date precipitation variability was fairly uniformly distributed, but recently Germany’s western regions have seen an abnormal level of variability. In other words, the precipitation in Germany’s western regions is less predictable.  

In the past, a lack of data and analytical tools made it impractical to effectively manage climate risk, forcing businesses to make less informed decisions. Gro’s climate indices provide daily insights that empower global businesses and governments to remove guesswork as they plan for and manage risks around climate change. We are very excited to expand our portfolio of climate analytics as we continue to focus on impacting the way the world approaches managing risks related to climate change. 


Gro is the story of what on earth is going on. We give our customers the ability to see the big picture and act on the small details where ecology meets economy.

Contact our sales team to schedule a demo and learn more today.

Get a demo of Gro
Talk to our enterprise sales team or walk through our platform