FORE-SCE Land Use Model with Terry Sohl

Hi, my name is Terry
Sohl, I'm a research

physical scientist.

ForeSce model is something
we developed here at EROS

and what we wanted to do
was move beyond the dates

for which remote sensing
data was available.

Where we wanted to be
able to take these USGS

produced land cover
databases that are

produced from Landsat
project them both

backward and forward in
time so groups could look

at how weather and climate
are impacted by land use

and visa versa. What's
happening currently and

then what could potentially
happen out into the future.

When you're using a land
cover model, and going out

into the future, you don't
know what those conditions

are going to do. So the
uncertainties are obviously

a lot higher. And so what we
do is we use scenario

frameworks. And so we're not
weather forecasters, we're

not trying to predict exactly
what's going to happen in

the future. What we do is
provide a suite of scenarios

that try to cover that
uncertainty.

And so in as area like the
great plains we can look

at recent trends, look at
what's happened with 

agricultural land recently,
and extrapolate those

trends out in time, and that
would provide one scenario

What we can also do is look
at alternative scenarios.

Economic scenarios change.
Biofuel use changed and

so those are the type of
impacts on regional land

use that are very hard to
predict. And so what we

try to do is model multiple
scenarios try to

represent that uncertainty.

Several years ago I published
a paper that looks at

the impacts of projected
land use and projected

climate on bird populations
across the US. And that's

something that's been done,
from a climate aspect, there's

a lot of different climate
models that are out there.

There's not a lot of land use
models, especially of the 

kind that we produce here
at EROS. And so we're

providing a unique product
that allows somebody like

a biologist to look at
not only how projected

changes in climate might
impact bird populations

in the future, but how
projected changes in 

land use might impact
bird populations.

The city of Minneapolis
gets most of their drinking

water from the Mississippi
river. In the last few years

they've noticed things
like cyano toxins and

other other products that
are an offshoot of enriched

water from agricultural runoff.
And the reason for that is land

use change. North of
Minneapolis there's a large

tract of forest land that was
bought by the largest potato

growing company in the
United States and they

converted a lot of this forest
land to potato production. 

And so what we're doing for
this is we're producing scenarios

of land use change in the
upper Mississippi river basin

that allows them to potentially
plan for what may happen for

their water treatment needs
out into the future.

We've completed land cover
projections for many scenarios

for the southern great plains
and it's a unique area from an

agricultural perspective. So a
good example is the panhandle

of Texas where it's one of the
United States major cotton

growing regions, and you
need irrigation to support

cotton in that pretty dried
landscape. And in recent

years the aquifer has declined
to the point where that

level of irrigation availability
is just not there. And so

you're starting to see some
potential shifts in

agricultural patterns. And there
are some unknowns with it.

Our scenario based framework
that we use allow us to cope

with those unknowns. So as
the irrigated water

availability declines there's
some question as to wether

they're going to move to
dryland crops like wheat, 

or whether a lot of that land
is going to revert to a grazing

regime with pasture and
cattle production.

And what we can do is
provide scenarios for both.

And that allows land planners
people looking at biodiversity,

water quality and all these
other potential impacts

of land use, to look at what
are some of these potential

future scenarios and how
we may adapt no matter

how the landscape changes.

The ForeSce model was
designed to take advantage 

the land use products that
are produced by USGS.

And so we can look at the
past. We can look at landscapes

that were mapped by Landsat
and we can look at landscape

patterns. And so when we model
things out into the future,

we are tapping into that
Landsat database so that

we're representing landscape
patterns that are regionally

realistic based on what's
happened in the past.

There really are no other
models that can model

broad scale land use across
broad regions or even at

a national scale that can
map the spacial resolution

of what we're doing using
actually land management

and ownership boundaries
from the USDA to

represent field by field
modeling and thematically.

We can all the land cover
classes found in the national

land cover database, and that's
what makes our model unique.

From a biodiversity standpoint,
from a climate perspective,

from a water quality standpoint,
just the wide variety of

applications that we're able to
support given that we're

trying to produce things that
are consistent with the land

cover databases that USGS
are already producing.