Daycare centers within poor air quality zones

In this exercise i analyzed the distribution of daycare centers within poor air quality zones. I used multiple data sources to do the analysis. The traffic volume data was a table created by the Helsinki urban environment division. Daycare center data was created by Information technology unit, the file format was .KML. The road network data we used was the Digiroad data made by Finnish transport agency and the air quality table was created by HSY.

After the initial planning and CRS transformations it was time to start building the models. I made the first model with QGIS. It took some time to understand how the QGIS model builder works, but once I got the hang of it, it turned out to be an easy and understandable method for modelbuilding. I already had some experience with the ArcMap model builder, so it didn’t require much thinking. The main difference between the ArcMap and QGIS models is that the QGIS model is “reusable” in the sense that it can be applied for a different set of data whereas the ArcMap model is made for a specific dataset. For example, you could use the model made with QGIS to do the same analysis for another city. In practice, however, the data would have to be very specifically refined in order to work similarly to the data used in this exercise. The straight-forward approach of the ArcMap model builder is probably more suitable for the PPDAC (Problem, Plan, Data, Analysis, Conclusions) framework. However, I think it’s mostly about personal preference and for me, the QGIS model builder worked better.

In the bonus task an additional buffer zone was created based on the recommended distances from roads. The first step was to create a new column for the distance variable. I used a basic IF- “chain” (if(”2016_1″<=5000,20,if(”2016_1”>5000 AND ”2016_1″<=10000,40…) etc.) to classify the recommended distances based on traffic amounts. After the classification the same variable distance buffer was created for the roads, but this time based on the recommended distances. To avoid overlapping of the daycare centers within the minimum and recommended distances, the overlap of the two buffer zones had to be erased. Rather amusingly, the QGIS difference-function did not work on the version that I used for this exercise, so I had to use the SAGA-version of the same function. Lastly the model extracted the daycare centers within the min-rec -difference buffer. (Figure 1)

Figure 1: Finalized model of the air quality zone – daycare unit analysis.

Now we finally had the results (Figure 2): 20 daycare centers within the minimum distance and 33 within the recommended distance. So in total 53 daycare centers are at risk and 964 are not. As the PPDAC model suggests, this leads to further questions: How is this possible? Were the daycare centers established before or after these air-quality zone classifications were published?

Figure 2: Distribution of daycare centers within poor air quality zones in the Helsinki area.

Työpaikkojen jakautumista pääkaupunkiseudulla hotspot-analyysein

Lähtöaineisto tässä harjoituksessa oli taulukkomuodossa, mutta se sisälsi myös koordinaattitietoa, jonka avulla se pystyttiin muuttamaan vektorimuotoiseksi pisteaineistoksi. Pistemuotoinen aineisto muutettiin edelleen rasterimuotoiseksi 500 x 500 metrin tilastoruudukoksi ja tästä tehtiin karttaesitys työpaikkojen alueellisesta vaihtelusta pääkaupunkiseudulla. (Kuva 1)

Kuva 1: Työpaikkojen levittäytyminen pääkaupunkiseudulla 500m ruudukossa.

Hotspot-analyysillä halutaan selvittää, muodostaako tarkasteltava ilmiö alueellisia klustereita. Se on siis eräänlainen geostatistinen työkalu. Itse hotspot-aineisto luotiin neljälle eri alalle. Hotspotit luotiin 1500 metrin tarkasteluetäisyydellä, eli jokaisesta pisteestä tarkasteltiin toimialan keskittymistä 1500 etäisyydellä. Klusteroitumisen tilastollisena merkitsevyystasona käytettiin viittä prosenttia. Kun hotspotit oli luotu kullekkin alalle, pisteet interpoloitiin IDW-menetelmällä. Nyt syntyi neljä karttaa, joissa näkyi kunkin alan työpaikkojen jakautuminen pääkaupunkiseudulla.

Kuva 2: Eri alojen hotspotien jakautumista pääkaupunkiseudulla.