Kind: captions
Language: en
hi everyone and welcome back to this
series on how to do
open source investigations from home i'm
ben
and this is part seven so let's get
started
[Music]
in this episode we're going to look at
how to create a time lapse with
satellite imagery
and a time lapse is essentially
a multiple amount of satellite images
over a specific string of time whether
that be over weeks or months or even
years
to identify changes on the earth and
this might be environmental changes or
it might be micro
changes like in a building or when a
building has been destroyed or when
fires have been through an area
or even urban expansion or ships as
we'll see
i've got a couple of examples on the
screen at the moment such as this
solar power plant which is one of the
biggest in the world and we can see
how those panels have changed every
single day
and another example from the port of
singapore where we can see the vessels
moving
backwards and forwards every single day
and we might even be able to identify
times when there is less trade
for example in the port we can do all of
this
by a few simple and easy to use
resources
that are completely free in the
description box below you'll find links
to all of these places just in case you
want to follow them up and replicate
these steps
as we go through them with the same sort
of case studies
but this is a solar power plant in
morocco it's one of the biggest
in the world it's actually three
different sections so we've got
this section of solar panels down here
this one
and then we have this one as well which
is quite fascinating because
if we go to google earth up here which
is the way that we can scroll back in
time
we can see here that the layout of the
solar panels is different to how they
were here
and if we keep going back we can see how
those may
have changed now you'll notice that when
i go back through
the google earth imagery that's
available here sometimes it jumps a
couple of weeks or a couple of months or
even a couple of years
and if we go back to 2016 we can see
that there were no solar panels
whatsoever
and if we click from 2018 we go ahead a
couple of months
we go ahead almost a year and so you get
the sort of idea of what we're looking
at with google earth imagery
but as we went through in one of the
previous videos we can do this
with much more imagery available at our
fingertips
so what i'm going to do is use sentinel
hub
to do that and sentinel hub is one of
the things that we went through
in one of the past videos where we use
that to identify
a window of time of one of the images
that we geolocated and videos that we
geolocated
just using that freely available
satellite imagery
so what i'm going to do just in advance
is i'm going to
open this image up in google maps
and i'm just going to pluck my
coordinates out of there just by doing a
hard press
on the location and we'll see my
coordinates pop up down here
and these are the coordinates that we
want to get our hands on
so i'll copy those and what i'll do
is i'll open up sentinel hub eo browser
you'll find a link
in the description below to sentinel hub
eo browser
and you will notice that it's a little
bit different to the sentinel hub
playground
that we used last time center ohio
playground is very
easy to use it's very user friendly
whereas
sentinel hub eo browser is much more
advanced and has a lot more different
features available to it
one of those features is creating a time
lapse to create a time lapse you do have
to have an account with sentinel hub
but that's completely free and you can
sign up right here just on the sign up
form and i'll provide a link to that as
well
once you open up the sentinel hub eo
browser and you log in you'll notice
that you're faced with something that
looks a little bit like google maps
and there's no satellite image review so
what we're going to do to fix that first
of all we'll go to that location that we
pasted in or that we copied in from
google maps before
now we can't really see the satellite
imagery of that because we have to do a
search
through sentinel's database over here of
landsat imagery or the sentinel 2
satellite system or modus or gibbs
system as well so we'll just use the
standard sentinel 2 satellite system
which is also what we're looking at in
sentinel held playground
and we can either select the panels over
here so these are satellite runs that we
can view
or you can just click visualize on the
left tab over there
now we're a little bit zoomed out at the
moment so we won't be able to see our
solar
our solar power plant too easily right
now so what i'm going to do
is zoom in and then i'll re-click on my
coordinates and it'll
it'll center my my view over those
coordinates
so we can see the solar panels now but
it's a little bit of a cloudy day
and when we make our time lapse we want
to make sure that we have as least
clouds as possible
because otherwise we're not really going
to see much
so this is a cloud-free day uh over the
solar power plant
and it looks quite good so i'll go ahead
and start making my time-lapse
the time-lapse is really easy to make so
all we have to do to create that time
lapse
is click on this icon over here on the
right
which says create time lapse animation
now we can use this to use this window
to
center over the area that we want to
feature in the middle of the time lapse
and i'll just click on that
so you'll notice that it's brought up a
whole new interface which allows us to
really filter down the archive of
satellite images and that's what we're
doing here we're trying to pick out the
images
that we want to put together in a in a
gif
or to create our time lapse and
all of this is basically filtering down
the archive so we can have
one image per orbit or day or week or
month or year
you'll notice sometimes that you can't
have too many images in one time lapse
so you might have to do one image per
week rather rather than one image per
day
for the date range of this i'm just
going to
uh because we saw before that it's it
was constructed in about 2016 2017
so i am going to take mine back to about
2018 here
and i'll just select january
and what we have down here this little
slider icon
is a way to restrict the cloudy days
that we don't
want so remember how we're looking at a
whole archive of satellite images here
what the sentinel hub platform has gone
ahead and done is to categorize all of
the clouds in that so that we can drag
this down
and start cutting out the percentage
rate or percentage value of images that
have a certain amount of clouds in them
i like to usually drop this down to
about 10
and then click on my search and run the
search for satellite images so i'll do
that now
this always does take a little bit of
time as you can imagine
it's searching a lot of images and we've
come up with that same
issue of the maximum number of images is
so what i can actually do as well is
just try and do this per week
and you'll notice that on the left side
here it's starting to bring up my images
and already you can see
how different they look with the solar
panels
in there facing their different
directions towards the sun
or in a different easterly or northern
direction depending on which way they're
trying to
trace that sunlight it does take some
time
to load all of these images because as
you can imagine this is a very
large database of satellite imagery
taken on every square
meter of the world so going through this
is a huge catalog
but it does load eventually so here we
are we have our
time lapse and what we have down here is
a way to alter the speed by frames
so if you can imagine this like a gif
that might be made of 10 or 20
images we want to speed this up so if i
click on play at one frame per second
you'll see how slow that is sometimes
this is useful
if we're just looking at a before and
after of perhaps a bombing or a fire
that might have happened over a week and
we're only looking at two different
images
it's a great way to do a comparison and
contrast
so in one of the previous episodes we
looked at a village that was burnt in
myanmar
and we had a look at a simple before and
after to see that change
for this one the effect we're going for
happens over a much longer period of
time
so i'm going to bring this up to a 10
speed so
10 frames per second and click play on
that
and already we can see how the solar
panels
move over time we can see some of the
directions that they might have or we
can see that
some are displayed and some are not and
this is a great case study to show
people
how time lapses actually work
so if i wanted to include this in a
digital report
or anything like that i would click
download and it would simply
download a gif for me to use
online or on a website or on any social
media platform
as well which is another quite useful
feature
of this so another example
uh of creating a time lapse uh for a
different cause is to
use for example the islands that are
popping up in the south china sea
uh so this one is called mischief reef
which has been quite heavily developed
over time so if we zoom in a little bit
closer here on google earth like over
the airstrip
for example and we go back to what this
island looked like in 2005 or
2012 even scroll through a little bit
more
you can see that in 2013 there was
no real development there was no vessels
around
go forward in 2015.
2016 there are a couple of vessels
around here and there's a
there's something going on like a like a
sand pumping
uh projects down here
go forward a bit more oh there's uh
definitely something popping up there
and also definitely up here
as well so we can see the difference
between this one
in 2015 in this one 2016
and then we can see that just develops
so much more and grow
so much more over time
so we can also visualize that
just using sentinel hub
so again i'll grab that those
coordinates
just through google maps and we can open
this up
in the eo browser which i've done here
and again just going through the
processes so what i'm doing is i'm
creating a search for the available
imagery in the area
i'm just clicking on this run to view it
so
visualize this one you can see that's
what we've got
so we knew that the
real production started in between
sort of 2015 and 2016.
now our sentinel 2
if we create a time lapse of this sort
of area
i don't think we'll be able to go back
that far so only 2016.
possibly not january so we also have
another a range of
other satellite imagery as well that we
could try
we do have landsat imagery
so we can try the landsat imagery um
that should go back a lot further so we
could try it for this one
and yeah so we've got 2013 we can do
landsat imagery for i might just go back
to march since that's the earliest one
it's probably going to have to make me
do on a weekly basis
also the the area around here we noticed
just before by looking around is quite
cloudy
at certain times the year so i might
just drop this right back to
10 and that should be fine for the the
week count
i'll speed this bit up uh so that we get
this
imagery loaded as quickly as possible
okay great so we've loaded our available
imagery
it seems to jump forward quite a few
days just because of the cloud coverage
you'll notice that even though i've
set my cloud coverage rate down to 10
there are some cloudy days but the 10
will cover the whole run
probably not just this image uh so those
independent
runs or satellite runs will be kind of
like that so i'll do the same thing as
what i did
last time i'll set this to a 10
frames per second speed and we can see
the growth of these islands
uh as they were built up and this is uh
this island is not the only one in the
south china sea to be built up like this
with a
a large airstrip and air force hangers
and a number of sustainable facilities
so it's it's good to check these things
out but this is a great sort of
visualization just to show
one of the many islands that have been
built up over time in the south china
sea
and again just like we did last time we
could download this
and and use this as a gif in any
reporting or anything like that
this also goes to show that we don't
just have to rely
on the traditional standard recommended
satellite imagery but for this one we
use landsat 8
it's always good to go through and make
yourself aware of
the type of imagery that's available
in sentinel hub eo browser it also
provides
a breakdown of what each image
or each each provider has with its
spatial resolution
but also its revisit time so the revisit
time of sentinel 2
is maximum 5 days to revisit the same
area so
it's a great one to create those time
lapses with
as well as as well as landsat as well
which is what we just used there
so the next one i'm going to do uh is a
fun one in the port of singapore for
those of you
that are aware of the port of singapore
you'll notice that there are a lot of
vessels all of the time
it's a very busy area for vessels
specifically cargo and trades in this
region so i thought it might be a nice
idea to use this as a case study or an
example of creating a
a image time lapse
so we'll go ahead and do that now so
what i'll do same process
view in google maps
and i think this area has quite a few
vessels in it all of the time
so copy those coordinates across into
sentinel hub
eo browser so we've got our port here
the same view that we always have with
google maps and we always start with in
central hub
eo browser for this one i'm going to
choose the
sentinel 1 system
i thought this would be quite
interesting to look at uh
to identify the vessels in the port
so i'll just visualize this one and
you'll see how different
uh this really looks um
so it's loading that run over there so i
might just do a research
on that and just make sure we get this
run
so you can see already uh how different
that is
in comparison to what it looks like if
we use
uh and i'll load the same location in
central heart playground
if we use just the the normal
sentinel 2 system we can see how clearly
we can see
those vessels in the harbour there i'm
just going to leave this on
the the normal banding that it comes
with but we can also have a look at
for instance different ways of viewing
the sentinel one imagery as well
you can see that there's some quite
interesting uh ways there or also the
enhanced visualization um which makes it
quite interesting
so i might even just leave it with the
enhanced visualization or the the urban
uh sar visualization as well
and create a time lapse out of this so
same process
as before we
just click this play button
and that seems like it has them all
quite well there i might just do from
onwards
we can already see when that loads uh
how good
that looks just with the vessels so i'll
speed that up as well
and let's check that one out
yeah so that looks great you can always
just some of the findings you can make
out of this for example you can see that
there is a constant number
of vessels always in this pocket here um
you know the traffic route is mainly
going through there we can see that
there's a lot of sort of constant number
of vessels but these are always on the
move whereas these ones seem to be
a more so remaining the same
if we go into further analysis of these
we could always have a look at the
difference for example between
vessels during earlier parts of 2018 and
2019 in comparison to when coronavirus
was at its peak and see if there's a
difference in trade or things like that
this is a great way to map out also
shipping routes and maybe if there's any
congestion lanes so you can see that
this sort of time lapse isn't just great
for open source investigations on a
human rights or journalism level but
also
on trade or business intelligence and
and corporate research as well
and again we can download that to
have it in a gif and and run that
through to
either social media but this was a much
more of a niche episode
uh for you on how to create a time lapse
i really appreciate the the feedback on
all of these
episodes and specifically the liking and
sharing of the content
uh it really keeps me motivated in
creating things like this
and if you do find this useful please do
share it around with
other colleagues or friends or other
professionals that you might have that
are looking for content like this
and i look forward to seeing you in the
next sessions
[Music]
you