Lesson video

Hello, my name is Mr. Conway.

I'm very pleased to have this opportunity to guide you through today's lesson.

The emphasis is gonna be very much on the use of geographic information systems, otherwise known as GIS.

So let's get started.

This lesson is linked to the units about tropical rainforests.

By the end of today's lesson, the intended outcome is that you'll be able to use GIS to visualise deforestation and analyse the causes of deforestation.

You'll be learning to use GIS techniques, which can be applied to lots of spatial data we use in geography.

So some of the learning may be new to you, but I'm here to help you along the way.

To help us achieve the outcome, we need to learn or remind ourselves about certain keywords.

The keywords for today's lesson are remote sensing, satellite imagery, and aerial imagery, and they're all closely linked.

Let's look at the definitions for each one.

Remote sensing is the collection of data at any distance, which has been reflected or remitted by objects, living things, or areas.

And the data can be collected on different wavelengths of the electromagnetic spectrum.

The observations can be carried out by satellite or aircraft, including drones.

The word "any" as in any distance, is doing some heavy lifting here because remote sensing is often thought of as the gathering of data from a long way away, but it can be close up to such as a few metres with a drone or recent imagery of comets or asteroids from a few kilometres away.

Such close range remote sensing is called close-range remote sensing.

The next two key words are types of remote sensing.

Both will be observing much the same thing that is selective parts of the electromagnetic spectrum.

So we're dealing with satellite imagery, which is any type of remote sensing in orbit or travelling across space and aerial imagery is that which is done by aircraft or drones or balloons.

When we see images from remote sensing, we're often seeing an imperceptible combination of satellite and aerial imagery, usually as we zoom in and out of an image.

But sometimes satellite and aerial will appear alongside each other depending on when the availability of imagery is for a particular place.

There are two learning cycles for this lesson, looking at how to use GIS to visualise and analyse deforestation.

So we're going to look at the first of these learning cycles now, which is how can GIS visualise deforestation? GIS can be used to map the observations that are made by remote sensing.

In this image about the different wavelengths on the electromagnetic spectrum, we see longer wavelengths to the left, including radio waves, microwaves, and infrared.

The spectrum for visible light sits in the middle of the range.

Then the shorter wavelengths are shown on the right, including ultraviolet x-rays and gamma rays.

Remote sensing uses various techniques and filters to detect all of these, including the all important spatiotemporal changes that take place.

As mentioned when introducing the keywords for this lesson, remote sensing includes satellite imagery and the satellite that's appeared is Landsat 8, which is a very important remote sensing satellite for geographical observations.

The work it does is very closely associated with Landsat 9, which was launched in 2021.

They remotely sense the entire earth surface after 30-meter resolution about once every two weeks, and they detect infrared visible and UV data right across the range indicated by the purple arrows.

Remote sensing also includes aerial imagery, which is carried up by aircraft, drones, and weather balloons.

From a geographical point of view, remote sensing is especially valuable for the investigation of land use change.

Here we see a rainforest GIS layer in green with two base maps.

First of all, imagery, then chartered territory map, but we can't see the GIS layer so we can if we make the chartered territory map opaque so that we can see the countries which have rainforests.

And in this lesson, we'll be looking at the causes of spatiotemporal changes in these selected countries with rainforests, particularly in the tasks.

So refer back to this map if you need to check their locations.

Remote sensing can make a very important contribution to research about deforestation.

In this example, we see how satellite imagery is being used to monitor impacts of different approaches to the way deforestation is managed in two different countries.

The animated geo GIF shows a small part of the Amazon rainforest near Acailandia in Brazil and along the border with Bolivia over a 25-year period, comparing the years 1999 and 2024.

We can see what appeared to be very different approaches to deforestation on the Brazilian side compared to the Bolivian side, which is still largely intact.

This may not mean that one country is more or less guilty or culpable than the other.

In recent years, deforestation has accelerated markedly in Bolivia, just not in this area yet.

Deforestation rarely happens in an organised large scale way.

Instead, it tends to happen in thousands of small steps, perhaps summed up by the saying, death by a thousand cuts.

It's exemplified by what we see on the Brazilian side of the border.

A road is built then deforestation begins to happen on either side of it.

This is Acri state highway 475.

Then smaller roads are built at right angles to the main road.

Further deforestation inevitably happens along these smaller roads as well, and a geographical pattern called fishbone deforestation develops.

Other roads are built, more fishbone deforestation patterns develop and they join up with other fishbone.

This causes the rainforest to undergo what's called ecosystem fragmentation.

Pieces of the rainforest often survive, but they're no longer joined up with other parts of the rainforest.

Consequently, habitats of flora and fauna are broken up.

Then species often go into rapid decline or even become extinct because they can no longer associate with each other and reproduce effectively.

Remote sensing can also support research about deforestation due to mining, which is having a major impact in several rainforest areas.

Here, we see the results of monitoring of gold mining in the Amazon rainforest.

One area where this is happening is in the Madre de Dios region of Peru.

This animation uses remote sensing to show the progression of rainforest loss over just five years between 2013 and 2018 with different colours for each year.

In the next image, we see the same area but at a larger scale.

One key issue with obtaining remote sensing data from rainforest is that it rains a lot.

So the areas are often covered in very thick clouds, but this image uses sophisticated synthetic aperture radar or SAR.

It can see through cloud cover at times when the rainforest is full of clouds and it's impossible to see through it.

That's always been a very difficult thing to do.

Then we can add the layer of earlier data to add value to the visualisation by showing the progression of deforestation, and that helps us to understand the human processes at work by identifying or confirming such temporal change.

Here's a swipe created in Esri Wayback, which you're going to use later to show how remote sensing imagery can be used to monitor the impact of deforestation due to gold mining in Peru over a 20-year period from 1999 to 2019.

Even though the earlier date looks like 2014 in the panel on the left, it's actually from 1999.

The gold mining here is what's known as semi-industrial mining, which is an overly polite term for what is often very poorly managed environmental exploitation of dubious or no legality, and it's clearly very unsustainable.

Almost all of the pale light brown scarring you can see on the landscape represents areas where gold mining has caused significant deforestation.

What does this look like close up? Let's take a look.

This is aerial imagery taken by a Peruvian military aircraft, shown the devastation caused by semi-industrial gold mining.

The ponds you can see are what are called prospecting pits.

They appear as thousands of tightly packed water filled basins.

They're often highly contaminated with mercury and sodium cyanide, which is used in the process of gold extraction.

NASA's earth observatory produced an excellent report about this semi-industrial mining.

Here are some short extracts, which I invite you to read for yourself alongside a closeup image of the prospecting pits showing a few dead tree trunks still standing.

Whereas deforestation can sometimes take place gradually in a haphazard way, perhaps illegally, there are examples where it's planned and approved by legislators.

Here, we can see this happening in Indonesia right now where an executive decision has been made at the highest level to build a new capital city in the rainforest.

The new capital is called Nusantara and it's located in East Kalimantan.

The goal is for it to be completed by 2045.

So the development is a major cause of deforestation in this area and it's happening fast.

We can see satellite imagery by Landsat 8 and 8 here showing less than two years of development between 2022 and 2024.

The earth is a big place, so analysis of vast amounts of remote sensing data is a huge challenge.

Consequently, significant efforts are being made to analyse it using automated methods.

So the data is often divided up into cells to make it more manageable.

Here, we've used the Esri Wayback app to see an example of a very small section of the rainforest in the area near Nusantara.

This imagery is from the worldview two satellite captured on the 10th of December, 2023.

The precise latitude longitude given in case you'd like to see that area yourself in Esri Wayback.

By using the cells, we can ask ourselves what land use is represented by the different colours and textures? Sometimes it's obvious but sometimes not.

For example, if these cells are fields, is the farming activity on them pastoral that's grazing animals or is it arable that's growing crops? We can see two areas of trees.

Are they the natural vegetation of rainforest or are they recently planted oil palms? We can see what appear to be roads.

Are we looking at a metaled road or a rough track? Sometimes filters can help us with this.

For example, the heat signature of a metaled road will be different to a rough track and infrared filters may be able to clarify this.

Analysis tools are being developed so that granular decisions like this can be automated.

They're using geospatial foundation models or GFMs with deep learning methods to classify each cell.

It's been said that GFMs are trying to do for satellite imagery what AI large language models are doing for text.

But they're still very much at the development stage.

They use deep learning methods called vision transformers to create embeddings, which are compressed representations of an image's key features.

Here's a GIS guide showing how to use remote sensing, historical aerial imagery and satellite imagery to visualise deforestation and its causes.

To do this, we're going to use the Esri World Imagery Wayback app.

It's an amazing collection of remote sensing imagery using both aerial and satellite images.

It goes back to about 2014 and sometimes, as we're gonna find out, even further back than that.

So if you go to the link provided, it should take you to this view here of Esri Wayback and it's been set up with the toggle swipe mode in place.

And what this means is that you can move the slider in the middle to compare one image with another from different years.

So you can use the panel on the left to select images that they have for this view and similarly on the right.

So typically, you would choose an older image perhaps and put that on the left and a newer image and put it on the right.

However, there is a word of caution with this, which is that the years and the dates they have here are not necessarily accurate.

This is something I only discovered back myself recently.

So although that says 2014, if you click on the image itself, it tells you that the photo was actually taken much earlier, the 15th of January, 1999.

Similarly, on the right, we see that this photo is listed as May, 2020, but if we click on the image, it tells us it was actually captured in 2019 on May the 15th.

So not much before, but it's not exactly 2020.

So we can use this tool to investigate the impact of semi-industrial or essentially illegal mining for gold, particularly in Peru.

So this is Peru.

If I just zoom out, you will see exactly where it is in the world.

So we're in the province called Madre de Dios in Peru.

You can see the area just there.

And the dark green colour is of course the Amazon rainforest.

And if we use the zoom control to look at a piece of the rainforest that's still intact, what we see the view from above either aerial or satellite image is we get this typical characteristic broccoli appearance of the rainforest.

Now if you pan and zoom around like that, you quite often end up getting a bit lost.

So you can either go back to the original link to find the original view or I'm gonna suggest you search for a place called Huepetuhe in Peru and there's only one place that is named that as far as I can make out.

So you should go straight to it.

So if we click that, it takes us to that region, which is pretty much the area where we started, but zoomed out a little bit more.

Now this is just one region of Peru where gold mining takes place.

And what we can see is just how vast this area is roughly from there across to there, this area here would be larger than the whole of greater London.

So it's a very big area, probably about 50 miles across.

So this is a major impact on the landscape and on the rainforest.

And we can use the zoom controls to take a close look at the area.

Do we see the canopy? Do we see that familiar bar broccoli effect? Instead of rainforest, we see a moonscape of gold prospecting pits.

The processes for extracting the gold use a lot of water and chemicals such as sodium cyanide and that flows or leeches into the landscape and is extremely harmful to wildlife and any humanity that happened to be in the area as well.

So it's a rather troubling scene of widespread devastation with prospecting pits all over the place, I suggest you explore those.

It's particularly clear in the more recent imagery, but if you zoom out, you can get some idea of the impact on the landscape by just going between the two dates.

And if you're not quite sure of the dates, remember what I said earlier on, you can check by clicking on the image to see when the photograph was actually taken.

So there are different ways of saving our work.

If we just zoom to an area and we think that looks particularly interesting to compare, this is similar to the area we had a look at to start off with.

We can see there's devastation to the rainforest by the gold mining in that area.

So what we can do is you can save the link to that by copying the link here where it's got a link symbol and there is another way which is to go to the toggle animation mode.

So if you click that, we'll see what happens.

What it's doing is it's offering you an animation to show all the images from all the years they have, but that can sometimes for reasons that we explained earlier on, can be a little bit misleading.

It's quite useful, you can see what's going on there.

But perhaps what you can do is just to choose just two of the images.

So I'm going to untick all the years except the years we were looking at.

So I'm unticking pretty much everything except I'm gonna choose the one from 1999, even though it says 2014 and the one that was from 2019, even though it says 2020.

So if we do that, we just get a comparison of two images, but it's quite flickery.

So what we can do is change the animation speed down to the lowest possible and then it's a little bit less flickery and easier to look at.

Then what you can do is download your animation and you can choose the dimensions of it.

Do you want a horizontal one or do you want a square one or a vertical one? It's got different dimensions there.

So you click any of those ones to download your image.

So perhaps we're gonna choose the high definition horizontal one and it's creating the MP4.

And when that's happened, we can download it to our device.

And it might be a good idea to give it a more meaningful name.

So in this case we call it Huepetuhe, Peru deforestation and our download has been successful.

Soon you'll have the opportunity to use remote sensing yourself, but let's just check up on some points from the video demonstration.

So the first check in Esri Wayback, which button is used to show historical imagery from two dates.

You may wish to pause the video here and restart it when you've selected your answer.

Well done if you selected toggle swipe mode, answer B.

Now for a second check.

In Esri Wayback, clicking on the aerial images can reveal popups with useful metadata about which attribute of the imagery.

Pause the video to have some thinking time.

Well done if you remembered that it's C, the exact date and year of the image.

Now for the tasks, you'll need to open Esri Wayback app on the link provided to take you straight to the area near wept away in Peru to investigate the impact of semi-industrial gold mining on deforestation.

So pause the video now to take some time to undertake the task.

And when you are ready, press play to obtain some feedback on these tasks.

Hopefully, you're able to undertake the tasks effectively.

Your swipe mode visualisation should look something like this.

We are looking at a recording of the swipe mode, but if you are able to use the download animation option, your MP4 file will toggle between the images that you ticked and it'll be similar to this.

If anything didn't correspond with that feedback, take another look at the video demonstration.

Our second learning cycle uses GIS to focus on the variety of causes of deforestation.

Over the last few decades, rainforests have increasingly attracted a wide range of economic activities, but the development management and sustainability of these activities is often contested and controversial.

There's been a very big increase in the number of people living in rainforest areas.

In Brazil, for example, a rapidly increasing population and internal migration has encouraged the development of large fast growing settlements in former rainforest areas such as Manaus, in Amazonas, Brazil, its population has quadrupled since 1970 and it's now approaching 3 million people.

Such cities will have largely secondary and tertiary activities, but economic activities in rainforest areas tend to be dominated by the primary sector such as extractive industries.

Here are some examples.

Rainforest climates happen to be well suited to commercial palm oil plantations in the primary sector.

This is due to a massive increase in demand for palm oil in the last few decades.

One key issue is that they replace highly biodiverse rainforest ecosystems with a palm monoculture, which has a major impact on local ecosystems leading to biodiversity loss and rapid decline of well-known species such as the orangutan.

As their name suggests, rainforests are very watery places.

So many rainforest areas have been lost to reservoirs built to harness water resources for transfer to drier parts of the country or to manufacture hydroelectric power.

A rapid increase in world demand for beef has led to deforestation to provide pastures for cattle ranching.

As we saw earlier, rainforest areas are often exploited due to their mineral resources.

These often like very close to the surface.

Apart from gold, rainforests have many other metals including bauxite, which is used to make aluminium and iron ore.

You can see here a massive iron ore mine in Carajas Brazil.

When the rainforest is cleared, the iron rich rocks show up rusty red due to the oxidation of the iron in them.

Rainforest contains species of hardwood trees, which are very much in demand in the construction industry.

Such forestry can use sustainable hardwood sources, but unfortunately, there is still a lot of illegal logging.

On more optimistic note, some countries have found that a much more sustainable way to earn income via their rainforest is to conserve them and look after them and develop ecotourism, which can bring in much needed income.

It can also support research as well.

There's increasing concern about deforestation by fire for rainforests.

This image shows remote sensing to monitor deforestation of cattle ranching near South Felix do Xingu in Peru, Brazil.

It shows images taken from 2013 to 2018.

Here's a closeup geo GIF of the imagery showing the progression of deforestation.

Fires would've been used a lot as part of the rainforest destruction here, and we can see one of them caught in the act as it were in 2017.

Fires have been used by humans to clear areas of rainforest for thousands of years, but this has now become unsustainable.

And wildfires have become an increasing cause of deforestation, very much exacerbated by climate change.

As you'll very soon, 3D GIS can be used to analyse economic activities which contribute to the causes of deforestation.

They can help us to see the geography happening.

Here, we see an example of deforestation by soy plantations in Mato Grosso, Brazil.

Our second video clip provides a guide demonstrating how to use 3D GIS to analyse economic activities contributing to the cause of deforestation.

This is a very brief GIS guide showing you how to use 3D GIS to analyse economic activities which may contribute to causes of deforestation.

The link provided will take you to a scene called Deforestation 3D.

This has been created in ArcGIS Online scene viewer.

You don't need to sign in.

You can if you want to, but you don't have to.

And then there are some tasks to perform based on some slides that have been saved in this presentation.

So to access those, there are two ways really.

You can go to slide manager, which opens up all the slides.

You can see here these are saved views of the 3D scene.

And you can click each one either here.

Alternatively, I can click on the thumbnails at the bottom of the screen to see these.

So if I click on one, you get the same result.

It zooms in on the same place that you would've seen if you'd click this slide here.

So you can use slide manager.

Alternatively, you can use presentation.

So you present the slide like this, and then you can click the arrows to go from one place to another.

There are eight views, all of them are interactive.

So for example here, you can actually move it around.

You can zoom in and out, but it's not going to lose the view.

'cause if you want to go back to it, you just go back one to the place in Borneo, we were looking at.

If you want to go back, you just literally click the arrow and it'll go back to the original view.

So it doesn't matter how much you alter the view, the slides will retain the same viewpoints.

So to perform your exploration and analysis of the economic impacts on rainforests, you can look around the 3D views to actually see the geography happening, but there's also very useful information in the captions underneath each slide title.

So you do well to read those carefully as well.

So I hope that's clear and that this will help you to investigate economic activities in the way that they have an impact on rainforests.

Soon you'll have the opportunity to use 3D GIS to analyse economic activities contributing to the causes of deforestation.

But let's just check on some points from the video demonstration first.

Which economic sector is dominant in rainforest areas? Pause the video here if you wish to have a think about that.

The correct choice is A, the primary sector which includes activities such as agriculture and mining.

Here's our second check.

Which of the following three economic activities you see here is likely to be the most sustainable for rainforests? Again, pause the video here if you wish.

In this case, the correct choice is C, ecotourism, which can provide a much more sustainable way to operate economic activities in rainforests.

Now for the task for which you'll need to access a ready-made 3D scene called deforestation 3D using the link provided.

In task one, you're going to use the 3D GIS to visualise and analyse economic activities in rainforest by using slides to see some saved views.

Then complete the table below using the slides and information in the slide captions.

In task two, you're going to consider a statement Deforestation in rainforest is caused by a range of economic activities.

For task 2A, try to decide quickly the extent to which you agree with it.

You can use the extent-o-meter provided.

A good way to think of this is to consider it on a scale of zero to 10, where zero means total disagreement and 10 means total agreement.

This is very rare of course, because opinions are rarely that categorical.

Most people will tend to have mixed views about issues where they might broadly agree.

So a mark between six and nine might be appropriate or they might broadly disagree with a mark between one and four.

Or of course, people may sit on the fence with a marker of five.

Once you've framed your position on the statement, the main part of the task is to write a plan to answer this question informed by evidence from your learning about deforestation using GIS.

So pause the video now to take some time to undertake the task.

When you're ready, press play to obtain some feedback.

Hopefully, you were able to undertake the task effectively.

For task one for the table, your 3D investigation should have informed your table as follows.

Indonesia is well known for palm oil plantations producing around 60% of world production.

Cattle ranching is a cause of deforestation in many countries, including Columbia.

Brazil is one place where deforestation is driven quite a lot by mineral exploitation such as iron ore mining, and that of course, is in the primary sector.

We looked at gold mining in Peru earlier, but this happens elsewhere, including in the DR Congo or Democratic Republic of Congo.

Deforestation is often due to logging for hardwood timber, such as in Madagascar.

The wateriness of rainforest leads to their exploitation for hydroelectric power, such as the Tucurui Dam in Brazil.

So the manufacturer of electricity means that the correct classification is secondary.

Ecotourism is a well-established sustainable economic activity in many countries with rainforests.

A good example being Costa Rica.

Ecotourism, of course, is in the tertiary sector because it provides a service to holiday makers, but it can also be classified as quaternary 'cause it also supports research and development.

And finally, as discussed earlier, Indonesia is building a brand new government settlement, a new capital city for the whole country.

So the services provided by the administration there will place their activity in the tertiary sector.

For task 2A, to what extent do you agree with this statement? You could have placed your X anywhere on the extent-o-meter, but here's an example for task 2B in which you needed to write a plan to answer this question.

You may have included points such as these in your plan.

It would be important to mention the wide range of unsustainable economic activities in many countries.

Make sure you're using the economic sectors to structure your thinking.

So have a look at these examples, the primary sector examples, the secondary sector, tertiary, quaternary, and the more sustainable rainforest activities such as ecotourism.

A plausible conclusion might be along these lines.

There are strong reasons to agree that deforestation in rainforest is not due to one or two factors, but a wide range of human activities, many of which appear to be unsustainable.

If your answers were very different or you recognise some errors, take another look back at the video demonstration and resources.

Well done.

What we've been doing in this lesson is learning some GIS skills and then applying them in different situations.

It's this kind of deliberate practise that can help make your GIS capabilities even more fluent.

Let's summarise the learning in this lesson with these key points.

GIS uses remote sensing to investigate spatial temporal changes in rainforest.

That means change over space and time.

Remote sensing includes both satellite imagery and aerial imagery.

Remote sensing can show change over time using historical satellite imagery and aerial imagery.

We can use swipe tools such as the Esri Wayback app or 3D GIS.

Remote sensing can be used very effectively to monitor and manage deforestation.

And finally, remote sensing can be used to analyse a range of economic activities, which can contribute in various ways to the causes of deforestation.

So we found out how to use some very effective GIS tools in this lesson.

Excellent work.

A good way to follow this up is to use the same methods perhaps in different locations.

Hopefully, you found this learning interesting and useful and I very much look forward to the opportunity of learning with you again.

So all the best and bye for now.