When I started reporting about climate change driving migration nearly two years ago, I immediately arrived at an unsettling realization:

The research pointed to a likely mass migration that could dwarf anything we've seen in millennia.

But I wanted to know more.

There was very little research on which people would move, from where, and in what numbers.

So we decided to do what we could to change that.

It turned out, of course, to be every bit as difficult as we were warned, and far more so than I imagined.

How we did it:

I found a geographer at Baruch College, Bryan Jones, who had forecasted for the World Bank how climate change would uproot people **within their own countries.**

His work found that as many as 143 million people would become climate migrants within their own borders

But, he hadn’t tried to apply the models to movement across international borders.

So, we hired Jones and set out to build a new model that would forecast both domestic and international climate disruption.

We started by looking at people moving from Central America and Mexico, and northward towards the United States.

(That makes it sound easy. It wasn't.)

All of this was used to consider a range of scenarios -- Jones called them "What ifs.”

Like:

-What would happen if global business stays the same, borders stay relatively open, but we don't do much to stop climate change and it gets HOT?

-What would happen if the U.S. shut its borders, reduced trade, economic development in poor countries slows down, and we still do little to fix climate change?

-What if we all stop emitting carbon dioxide right now and climate change is slowed to a manageable crawl?

The models crunched all of these scenarios, but did it using lots of data that had to be extrapolated and its gaps filled in.

There was plenty of published, peer reviewed research to support those assumptions, but it had to rely on assumptions nonetheless.

...Scientists cautioned us not to put too much stock in the numbers, so we tried not to focus on them or pretend we had all the answers.

The model's real value, it turns out, wasn’t the numbers it produced. It was this:

Our model showed what could happen depending on how governments did or didn’t prepare for the impact of climate change.

It told us:

1. More warming will lead to more migration, both domestic and international. This point was unambiguous.

2. Open borders and economic development would lead to more migrants, but also probably safer and happier people, perhaps with less conflict and suffering.

3. Limited migration and stingy aid would trap large numbers of people in the places where they will inevitably run short of food and water, leading to greater suffering and even conflict.

But most of all, it told us:

The actions governments take now -- to curb emissions and to support economic growth abroad -- will make a world of difference, and will, without a doubt, affect millions of human beings. https://features.propublica.org/climate-migration/model-how-climate-refugees-move-across-continents/

This is part 1 of a series of 3. Our next piece will focus on migration within the United States. Sign up to get that story (and the big investigations we publish until then) here: https://go.propublica.org/bigstory-social