Month: March 2021

Here we are… the last lesson – what a journey it has been. It feels like I have learned so much during the past weeks and still I know that we have but scraped the mere surface of GIS. Still I think that we have covered a lot of basics and during this week I did feel quite comfortable combining bits of knowledge from earlier weeks as well as dipping into completely new functions and mechanics!

This week didn’t have as clear a theme as earlier weeks and was really free from. One was to choose a set of data and to visualize it to the best of one’s ability. I chose to make two maps focusing first on terrorism and refugees and secondly on piracy around the world. I chose these since I am really interested in geopolitics and the geographic distribution of various criminal activities. (I suggest reading two books: “Prisoners of Geography” and “McMafia”.)

So firstly, I chose to look at terrorism and refugees specifically in the regions of Europe, The Middle East and parts of Africa. I also focused on the years of 2014 and 2015 to especially see how to outburst of civil war in Syria and the emergence of ISIS would be seen in the data. The results are twofold and mostly hang on the definition of terrorism. The Global Terrorism Database (source for my data about terrorism) defines terrorism as “The threatened or actual use of illegal force and violence by a non-state actor to attain a political, economic, religious, or social goal through fear, coercion, or intimidation”. Since this definition underlines that terrorism is tied explicitly to “non-state actors” the data may not mirror the reality very well. Especially in regions like Iraq or Afghanistan the line between state- and non-state actors can be quite blurred.  According to both The New York Times (2014) and The Independent (2015) the vast majority of the leaders within ISIS are former members of Saddam Hussein’s Ba’ath government. Twenty years ago, these members were so called “state actors” and now they would be called “terrorists”.  On the other hand (as can be seen on the map (picture 1.)) the number of fatalities due to terrorism in Syria is lower than Iraq although a bloody civil war has raged there as well amounting a death toll of about half a million lives. In a civil war as complex as Syria’s with over a dozen military groups cross fighting each other someone (usually the loser) is bound to end up being labeled a terrorist.

Anyhow, as I said the results of my map are twofold. On one hand you can se a correlation between the phenomena, but on the other hand there are certain regions that do not agree with this correlation. For example, we can see the correlation between terrorism and refugees quite clearly in countries like Afghanistan, Iraq, Ukraine and South Sudan. On the other hand, there are many African countries not having very many fatalities due to terrorism, but still holding a high percentage of the population as refugees. Here the reason for refuge might be something else like drought for example. An interesting observation is that all the countries of former Yugoslavia holds quite high numbers of refugees. I am unclear as to what is the reason behind this. It is to be made clear that this data tells about the proportion of each country’s population that are refugees and not how many refugees that country has given refuge to. This is why for example Turkey doesn’t hold very high numbers on this variable. Another interesting case is the case of France. Here we see quite many fatalities due to terrorism. This is of course due to the 13 November 2015 Paris attacks.

Picture 1. The map shows fatalities due to Terrorism in terrorist attacks with at least a 100 injured or killed and the proportion of the population that are refugees. Global Terrorism Database (2021), Gapminder (2021).

On creating this map I must say it was quite a task. Firstly the data wasn’t in the right format so I had to fix it using excel and so forth. I think Sara Korpi (Korpi 2021) would agree with me as she seems to have had similar struggles stating that “the handling of data has been the most challenging part of this course…” (“kaikista vaikein asia tällä kurssilla on ollut tilastotieteellinen puoli.” ) Another challenge was that I had forgot how to use certain tools within Q-GiS, layers didn’t work as they should and so on… Well I must say that although challenging it was very satisfying and meaningful as I had go back to earlier week’s exercise guidelines to look at how we did certain things. After doing this it really felt like I mastered these actions. The old saying is true; repetition is key! Reading Lotta Puodinketos blog (Puodinketo 2021) I got the idea of putting the percentage numbers within the dots displaying the phenomena of refugees much clearer.  Boy did I struggle with getting this one right!

The other map I created displays piracy around the world has a temporal heatmap about piracy between the years of 1980 and 2021. I thought it would be cool to make a temporal map and it was easy enough to find the instructions on how to do this on the internet. It’s to be said that this is not about digital piracy, but good classic pirates! (“Arrr!” & “I am the captain now!”)

Picture 2. The map is a temporal heatmap displaying piracy around the world between 1980 and 2020.  NATIONAL GEOSPATIAL-INTELLIGENCE AGENCY (2021).

So what can we get from this map? It’s clear that piracy isn’t equally distributed around the world and that during this interwall spanning 40 years it’s more or less the same places popping up over and over again. Generally it can be said that acts of piracy are conducted manly in the waters around the equator and close to shore. The key regions in Asia are the South China Sea, the Indonesian Archipelago and the Bay of Bengal. In the Middle East piracy can be seen in the Persian Gulf. In Africa there are two main regions. The first can be found on the eastern coast around the Horn of Africa and the second can be found on along the shores of the former Slave Coast in West Africa. In the Western Hemisphere piracy can be found along both the entire eastern and western coast of South America excluding Chile and Argentina, as well as in the Caribbean. It is to be said that piracy, although to a much lesser degree can also be seen in the Mediterranean and even in the English channel one flare up can be seen in 1998.

As for looking at how the prevalence of piracy has developed since 1980 it can be said that it seems to come and go in waves of about five years. For example the reports of piracy in 1980 is quite low but in the latter half of the 80’s it increases to then again be seen less in the early 90’s. The absolute top of piracy during this period (1980-2021) can be seen in the latter half of the 90’s. Piracy seems to flare up around the world during 1996 to 1998 and then calm down in the early 2000. After this piracy seems to generally decrease significantly in the Western Hemisphere while it stays on the same level in Africa, The Middle East and South Asia. Since 2015 piracy has been especially prevalent at the shores of Nigera in West Central Africa.

It’s interesting to see that the prevalence of piracy seems to follow some kind of pattern with lows and highs in intervals of about five years. I don’t have data on this, but maybe it could have something to do with macro economic cycles that follow a similar pattern? Just a thought.

Finally I want to say some general words about this course. I think it has been by far the most fun and meaningful course I’ve had during this entire year and it really felt like you learned something new and useful. I really like studying GIS because it gives you a practical skill and not only theoretical knowledge. I think the format of this course worked splendidly as well, especially the recorded lessons were to much help. I also want to throw out a special shout out to all who’s blogs I’ve read, got inspired by and used as sources for my own blog! Here’s a shout out to you! And then I want to thank our Arttu Paarlahti for being a terrific teacher, being very flexible with his teaching and answering email very quickly! Thank you!

So that was it. I’m looking forward to future GIS-courses!

Till then… Cheers!

Alexander Engelhardt

Sources: 

Puodinketo L. (2021) 7. kurssikerta: Omaa työskentelyä. https://blogs.helsinki.fi/lottapuo/

Global Terrorism Database (2021): https://www.start.umd.edu/gtd/search/Results.aspx?chart=regions&casualties_type=b&casualties_max=&start_yearonly=2014&end_yearonly=2015&criterion1=yes&criterion2=yes&dtp2=some&success=no&region=10&count=100

Gapminder (2021) : https://www.gapminder.org/data/

NATIONAL GEOSPATIAL-INTELLIGENCE AGENCY (2021): https://msi.nga.mil/Piracy

Natural Earth (2021) :  https://www.naturalearthdata.com/

ArcGIS (2021): https://hub.arcgis.com/datasets/2b93b06dc0dc4e809d3c8db5cb96ba69_0/data

The Independent (2015):  https://www.independent.co.uk/news/world/middle-east/how-saddam-hussein-s-former-military-officers-and-spies-are-controlling-isis-10156610.html

The New York Times (2014): https://www.nytimes.com/2014/08/28/world/middleeast/army-know-how-seen-as-factor-in-isis-successes.html

 

This week I yet again skipped the lesson, but this time it really bugs me. Had I known that we were supposed to go outside to create our own data I would absolutely have participated, but since I learned about it only after the lesson I sadly didn’t. I had checked the news page “uutiset” on our moodle but there wasn’t anything there so I thought everything would run as normal. Anyways I think this week’s exercises have been by far the most interesting yet! We have used interpolation methods as well as looked at certain hazards.

To begin with I want to repeat my humblest apologies for not participating and contributing to the Epicollect-data. I thank you who provided us with it so that we all could use it for our own assignments. So, the idea was we were going to interpolate ordinal data about safety, attractiveness and pleasantness and to then layer traffic roads and buildings on that to see if there are any correlations. Interpolation is a method in which for example with raster data where some of the raster lacks data but surrounding raster all have data assigned to them the value of the raster within is filled with the mean value of the surrounding raster. This is a method that is very widely used across many different fields and is by no means specific for GIS. For example, if you want to make a picture sharper one can use interpolation to imitate a higher pixel count than the picture originally had.

As I was saying, we looked at correlations between settlement and safety, attractiveness and pleasantness. One thing that struck me about the data was that it was very clustered. There was about seven clusters where all observations had been made. Six of these where within the Greater Helsinki region and one could be found in Nummela. General correlations that I could see was that heavy traffic correlated with less attractiveness and less pleasantness while green areas where seen as more attractive and more pleasant. When it comes to safety there where some food for thought in my opinion. For example, a few places in the park between Northern Haaga and Maunala was generally seen as safer than a larger highway nearby. I wonder if the observations would say the same if it was nighttime?

For my interpolations assignment I chose to focus on the attractiveness of the Viiki area. This area had a good varied set of observations for an interesting analysis. There is one major highway in the north and northwest, Lahdenväylä and then larger and smaller roads intersecting within the selected area. The general picture is that the highway and the largest roads holds the most negative numbers for attractiveness while observations made further from these places posts generally positive reviews. However, there are anomalies in the observations and the results can’t be described in as a single correlation, nor do I have the proper insight into what it was about these certain places that made them divert from the general rule. For that an interview style set of data would need to be included explaining the anomalies. The results of the interpolation method can be seen below in picture 1. Blue colours display maximum attractiveness, while green displays a bit less of that. Yellow displays places that are rather neutral while orange displays rather unattractive places. Red is the colour for maximum unattractiveness.

Picture 1. The map shows interpolated data about the attractiveness of certain places.

The other part of this lessons theme was to look at some data about earthquakes, volcanoes and craters from meteoroids and to visualize this through maps. This is an interesting hazard to look at because there is an overload of quite specific data from decades past. I downloaded data and then reformatted it in Excel so that QGIS could make sense of it. This wasn’t to difficult as we have done similar things earlier in the course. That’s a good feeling; being able to use something previously learned to make some parts of the exercise much easier!

Firstly, I made two maps displaying earthquakes with different magnitudes. Picture 2. displays all earthquakes between 1980 and 2021 with a magnitude above 5 on the Richter scale. I chose to further display the magnitude of each earthquake with a colour gradient shifting from blue for the lightest earthquakes to green and yellow for earthquakes with a magnitude of 6-8. Earthquakes with a magnitude over 8 are displayed as orange and over 9 gets the colour red. A problem that I encountered here was that the absolute most common “blue” earthquakes totally eclipsed the other colours making the map not very intuitive as it would seem there are no heavier earthquakes. This would especially be a problem as it is often the heaviest earthquakes that make the greatest impact on human settlement while lighter earthquakes seldom post any risk at all. I couldn’t find a function that would select in which order the observations where supposed to be shown so to get around this I made a second layer only containing earthquakes with a magnitude of 8 and higher and assigned the appropriate classification and colours to these. This way one can see the orange and red dots much clearer. Reading Ilari Leino’s blog (Leino 2021) I saw a different approach to visualizing the difference in magnitude. Rather than colour he used size. In my opinion this was better and more intuitive way to display the phenomenon. This can be seen in picture 2. To put even greater focus on the earthquakes with the highest magnitude I chose to quickly assemble a map focusing on only earthquakes with a magnitude of 8 and higher. With this limitation the number of earthquakes is in the tens instead of 10 000. This can be seen in picture 3.

Picture 2. The map shows all earthquakes with a magnitude of 5 and up between 1980 and 2021.

Picture 3.  The map shows all earthquakes with a magnitude of 8  and up between 1980 and 2021.

Next up I looked at volcanoes. In a similar process I downloaded data about the location of volcanoes and then layered this onto my pre-existing project about earthquakes. Although I already new that volcanoes and earthquakes often appear in the same regions it was interesting to see this on a map of your own making. The reason to this is of course that both earthquakes and volcanoes appear in seismically active zones – that is zones between tectonic plates. In picture 4. you can see the map I created. It visualizes earthquakes as black dots and volcanoes as red diamonds. I noticed that Sanna Janttunen (Janttunen 2021) had made a similar map comparing the location of volcanoes and earthquakes but she zoomed in on region of Central America. This map was much prettier and I think it was a good idea to look at a smaller region for this exercise.

Picture 4. The picture shows earthquakes and volcanoes. As can be seen this mostly appear in the same regions.

So, by this time I thought I was pretty much done with the week’s exercises, but I decided to continue to play with the data. My idea was to somehow combine interpolation and earthquakes. After trying this I can say it isn’t advisable at the macro-scale we were at. Since earthquakes are largely concentrated to the border zones of tectonic plates there is really no point to looking at how interpolation affects this phenomenon. With this in mind I decided to instead make a clear example of how NOT to visualize something as this is also an important lesson to learn. Anyhow it was just interesting to examine how the interpolation tool functions and fun to look at the ridiculous results! Picture 5. displays interpolated data about the magnitude of earthquakes. This got me thinking though that surely there must be some use to interpolation conserning earthquakes and looking at a single earthquake the picture is very different. Often when measuring the impact of an earthquake there are certain specific stations at which the effect of the earthquake is measured. By using interpolation one can estimate the effect between the stations without actually measuring it for every single spot. It’s interesting to see that interpolation could be widely used at this scale while it is totally nonsensical at a global scale. Picture 6. shows a map which uses interpolation to estimate the impact of an earthquake in the different zones.

Picture 5.  Interpolation of earthquakes with magnitude taken into consideration. Totally useless really. Still one can see resemblances of the borders of the tectonic plates.

Picture 6. The map shows a 3D interpolated model of seismological magnitude data for the Baradighi Tea Garden in the Himalayas. (Source: Biswas, Mery & Paul, Ankita. (2020).)

 

Sources:

Leino I. (2021) Kuudes kurssikerta. < https://blogs.helsinki.fi/ilarilei/ > 01.03.2021

Janttunen S. (2021) Hasardihommia < https://blogs.helsinki.fi/smjantun/ > 01.03.2021

Biswas, Mery & Paul, Ankita. (2020). Application of geomorphic indices to Address the foreland Himalayan tectonics and landform deformation- Matiali-Chalsa- Baradighi recess, West Bengal, India. Quaternary International. 10.1016/j.quaint.2020.12.012.