Lecture 1 – Introduction to the course
Week 1 was an introduction to the course and brief overview of what topics would be covered during the rest of the semester. Most notably, the importance of taking a geographical perspective in analyzing socio-economic problems such as crime analysis, and for instance finding a correlation between lead levels in gas and crime rates.
The three themes and areas laid out for the rest of the semester were:
- Landscape Ecology
- Health Geography
- Crime Analysis
We looked at the 5Ps which are fundational in geographic processes and which should thus keep in mind when carrying analyses. All these being mutually dependent and affecting each other.
- Processes
- Patterns
- Place
- People
- Perspective
Taking a geographic perspective on events and patterns that occur is very important in all disciplines, and patterns themselves can change depending on the respective perspectives that can be taken on a particular issue. And keeping in mind the purpose of the analyses in regards to decision making, and to whom the results are going to be shared is key in choosing which perspective to approach a problem from. In fact, context is of the utmost importance in spatial processes, and it is very important in analysis patterns. Indeed, for instance, understanding the temporal, spatial, socio, economic and political context of an environment is critical when carrying an analysis and deriving conclusions; key points which we should always keep in mind when working on our labs and writing our reports.
Lecture 2 – Why GIS is important
So why is geography important? This class was focused on explaining how and why geography is obviously important and a driving factor in patterns and issues we often wish to analyze and understand. Indeed, it is important to consider geography and the impact of geography on analysis; geography needs to be considered for meaningfulness and has a big impact on the results and interpretation of analyses.
When carrying out geographical spatial analysis it is important to choose an appropriate scale, however. Different scales and analysis of data and patterns at different levels can lead to completely different conclusions – for example, analysis crime data at a national scale can lead to very different outcomes than analyzing crime data around Sky trains. Moreover, considering an appropriate time frame and paying attention to organizational structures are also important.
Phenomena tend to have spatial dependence and spatial autocorrelation is also a significant concept and important process whereby patterns and occurrences of different events tend to cluster together in space. The lecture mentioned some issues in carrying out geographical analysis such as the MAUP (Modifiable Unit Problem) regarding issues in spatially aggregated data, as well as the scale effect, regarding issues of ecological fallacy. Finally, not only was scale mentioned as being important but the choice of units also reveals to be a crucial decision- making step in spatial analysis.
Lecture 3 – Understanding Landscape Metrics: Patterns and Processes
Lecture 3 focussed on landscape ecology and how GIS can be used to analyze interactions between spatial patterns and ecological processes, in respective landscapes. By using appropriate technology it is possible to analyze structurally and quantify patterns. We learned and reviewed during the lecture that through the field of landscape ecology it is possible to examine how processes shape patterns present in the landscape, as well as possible to analyze how processes can be shaped and transformed by the patterns. We also mentioned differences between stationary process, first order, and second order processes. First order processes whereby patterns are a response to environmental factors, and second order whilst patterns are changed and shaped by interactions not only with the landscape but with other objects as well. Brian also mentioned the potential impacts that climate topography, biotic and abiotic interactions, and natural disasters can have on those very processes and patterns. We also mentioned the importance in the presence of keystone species that have profound impacts on creating and sustaining spatial patterns and heterogeneity in landscapes. We mentioned once again the importance of scale in understanding how processes operate and in structuring the type of analyses constructed. Spatial configuration and the extent of the study area all have big impacts on the outcomes and interpretation of results. It is important to take into consideration when conducting GIS spatial analysis and important to keep in mind as well when taking into consideration the scale and context of the patterns and processes which are themselves being analyzed. Analysis of patterns through software and statistics such as fragtals and fragstats were briefly mentioned, and we also discussed the use of Fragstat in our lab to understand the evolution of landscape use in Edmonton, by particularly taking into consideration the impact of edges on analysing and concluding land changes due to uneven spread of patterns in the landscape.
Lecture 4 – Statistics: A review
Lecture 4 was more focussed on how to use data and use statistical analysis and regression models. Indeed, statistical models can be useful in summarizing data and in looking for patterns as well as correlation. It can help construct mathematically logical conclusions and hypothesis in regards to patterns and processes in the landscape of interest. The 4 different types of data were reviewed from our previous classes in GIS: nominal, ordinal, interval and ratio.
We saw that the data analyzed tends to be summarised in terms of dispersal from center points. Other conclusions can be derived from calculated correlation coefficients and r^2 values. Statistics in the scope of GIS can be used to measure the spread of data and the different relationship within and between different variable, and the results are usually portrayed through graphs and charts, which we have been able to experience thanks to our labs in the course. As part of our GIS journey and through our projects we have been able to understand the significance of regression models in understanding how variables impact events and patterns spatially; common regression models, mentioned in the lecture, and covered in the labs included ordinary least squares and geographically weighted regression models. A more complete evaluation of these techniques is reviewed under LAB 3. Other important outputs from such statistical analysis mentioned were p values, z values, f values and Moran I’s which provide explanations about the relationship found in the data: is it random or not random? Moreover, is there spatial correlation: is the data spatially clustered in space? These are important when carrying spatial analyses and in order to understand whether or not there are occurrences of spatial autocorrelation in the patterns and processes that are being analysed. Statistics are thus a useful tool in analyzing human and environmental processes both spatially and temporally.
Lecture 5 – What is Health Geography
Lecture 5 was about health and medical geography and how GIS and geographical perspectives can be applied to analyses and methods in explaining health, disease and healthcare processes. We also talked about the difference between health and medical geography, although also acknowledge how they are also ultimately linked to each other and to natural and urban environments, as well as to the social. Health geography is divided into 3 themes, and also was briefly mentioned to be connected to environmental justice.
1 – Disease ecology
2 – Healthcare delivery
3 – Environment and health
We also discussed the 5 approaches and perspectives and assumptions from the 5 strands of medical geography, going beyond medical focusses:
1) Spatial patterning of disease and health
2) Spatial patterning of service provision
3) Humanistic approaches to Medical geography
4) Structuralist/materialist/critical approaches to Medical geography
5) Cultural approaches to Medical geography
The field of medical geography has become so precise and broad that it’s name is arguably no longer suitable, and many analysis within that field are actually going beyond medical notions. Health geography has become a more integrated approach to explain medical and health distribution phenomenon across environments, as unequal distribution and spatial influences can be and are most often significant on the impact and spread of medico-social variables, patterns and processes. It was interesting to analyse further health geography since one of the first examples which I was given when starting to study GIS was John’s Snow mapping of Cholera in London in 1854. Consequently, this showed the great impact that GIS can have on medical discoveries and health analysis worldwide by focusing on spatial and temporal interactions with multiple variables, whereby statistical analysis as we saw in previous lectures can be applied.
Lecture 6 – GIS in Health Geography
Lecture 6 was a continuation of lecture 5’s introduction of important concepts in regards to medical and health geography. The lecture focused on major applications of GIS analysis to health analysis:
1- Environmental hazards
2- Modelling health services
3- Identifying health inequalities
4 – Spatial epidemiology
We defined spatial epidemiology as “the study of the distribution and determinants of health and disease-related states in populations, and the application of this study to control health problems”. Consequently, we focussed on that major area the most and discussed on how locations matter to the distribution of health and diseases. We looked at case studies and talked about the implications of using different scales and how that can lead to spatial misalignment and uncertainty in the analysis. We also questioned the tradeoff between precision and stability.
Lecture 7 – Is crime related to geography?
Mapping is an important part of GIS, and mapping crime can have a significant impact on reducing crime and improving the safety of citizens. But Yet an interesting question arises: Is crime related to geography? The lecture then focused on environmental criminology. Beyond talking about case examples, we also covered the three main theories behind environmental criminology, based on human behavior and distribution. Environmental criminology assumes that crime is impacted by our urban environment and thus mapping the environment is of notable importance.
- Routine Activity
- Rational Choice
- Criminal Pattern Theory
Using case studies it was possible to indeed point out how important GIS is in analyzing crime, the two disciplines are strongly interconnected. The four different levels of applications between GIS and crime analysis mentioned in class were the following:
- Intelligence and criminal investigative analysis
- Tactical crime analysis
- Strategic crime analysis
- Administrative crime analysis
We covered a case study of Vancouver’s crime and discussed the work of Dr. Kim Rossmo on the trail of Robert Pickton, as well as the challenges and limits in publishing spatial statistics of crime and about the issue of privacy. We saw that GIS can be a powerful tool in crime analysis as environmental criminology analyses the spread of crimes and criminals across areas of interest. Indeed, GIS can be very useful in mapping the distribution of crimes, and dispersal of offenders – by using GIS crime analysis can incorporate features of the urban landscape such as roads, Skytrain stations, downtown areas, and back alleys… For instance, tools like centrifuge and maptitude were brought up for solutions to low cost crime analysis.Using GIS is a way for analyst to combine both socioeconomic and geographic variables in order to understand crime patterns.
Lecture 8 – Use of GIS by Fire Departments
Lecture 8 focussed on how the Calgary Fire Department uses GIS to improve responses to fire hazards, increased public safety and reduce the loss of life and property. We saw that GIS was used by the city’s fire department to map risk, predict and visualize, map responses, and to help decision making, planning, and station relocations. Mapping and analysis required the use of Risk data and incorporated variables such as roadways, the wideness of roads, gas locations, building structures, historical events and response times of previous events.
In this case, thus, GIS was used by the fire department to analyze fire risks, distributions, and responses in order to increase the efficiency of future responses. GIS shows to be a useful tool for policy making, improving existing systems of operations and improving future decision making. Through the use of GIS, it was also possible for the fire department to locate future potential new locations for fire stations that would be optimal given the risk data used in order to provide response times between 6 to 10 minutes.