The images on this page can be used to broadly understand the general structure of our professional network. For example, we see strong country clusters but we also see many strong connections between clusters. This will not be surprising to anyone in the Pacific—oftentimes our first conversation with a stranger is ‘who do we both know?’ to establish mutual friends and colleagues. In the first network map below, you can see broad trends of connection and centrality within the network. If you download the high resolution version of the second map below, you will be able to see individuals’ names and their connections to others. Exploring these maps, you can begin to look for connections that your friends and colleagues have. When considering future collaborations, you can explore who knows who within the region.
For example, to find potential collaborators by country:
Follow the links to the second high-resolution map link, and download the “Original” high resolution size (may take a while to load)
Zoom in on network clusters from the target country, identified by the country color code
Find the contacts with the greatest centrality
Search their connections to find colleagues you have in common
This map represents the full network findings of the Pacific RISA Network Analysis project. From December 2012 to April 2013, 331 climate change professionals from Hawaii and the U.S.-Affiliated Pacific Islands region, including our partners across the Pacific Islands region, the United States Mainland, and other international partners, completed a network survey. These 331 professionals noted connections with a total of 967 people that they communicate with regarding climate, weather, and environment issues. This map uses a layout algorithm called Fructerman Reingold, available in the Gephi Network Analysis software. This layout algorithm automatically brings people who are highly connected in the network to the center, and people with fewer connections fan out around the perimeter to create a circular web. This makes it easy to see how many people are centrally connected and how well they are connected to each other. The size of each circle indicates the number of connections each person has and people with the most connections to each other are pulled toward each other in the center of the map. The circles’ color indicates each person’s country or region.
What can we learn from it?
From this map, it is easy to see that the people who are very highly connected in the network have many connections to a) each other, and b) hundreds of other professionals in the region. Around the perimeter of the graph it is easy to see where central people from a country or region connect with others from their country or region. In the center of the graph, the apparent clusters of color (e.g., the clusters of light blue circles, orange circles, and dark purple circles) indicate that people in each country or region are strongly interconnected. However, the fact that the clusters are so close to each other, and overlap with each other, indicates that strong connections in our region transcend locational boundaries.
This map is composed of the exact same people and connections as the map above. This map uses a layout algorithm called Force Atlas 2, also available in the Gephi Network Analysis software. This layout pulls clusters apart so that it is easier to see who is connected with whom in a cluster, and how well the clusters are connected with each other. Again, the circles are sized according to connectedness and colored according to country or region.
What can we learn from it?
What was hinted at in the Fructerman Reingold map (Full Network Map 1, see above) becomes very clear here—our professional network has strong country clusters. It is also quite apparent that these clusters are very highly interconnected. There is no single person or small group of people, for example, linking American Sāmoa (orange) with the rest of the region. Our network is highly connected across spatial boundaries—we are a series of islands connected in our professions.
Who are the most central people on these maps?
Centrality can be measured in different ways; each measure means something different. The following table lists the five “most central” people in the full network from each region, according to different measures of centrality. See below for definitions.
Degree centrality is the number of people you have listed connections to in this region. It is measured from 0 (no connections) to the network population minus one.
Eigenvector centrality looks at a person’s position within the network, basically measuring each person’s centrality according to the centrality of their connections. It is measured from 0 to 1 (highest eigenvector centrality in the network).
Closeness centrality is the inverse of farness, which is the sum of how many hops one must make to connect to all others in the network. Our network is pretty interconnected, so closeness centrality in this network ranges from 2.03 to 5.33 (and more than 2/3 of our network has a closeness centrality of 3.5 or lower).
Betweenness centrality considers how many shortest paths between pairs of people across the network pass through a given person. Oftentimes, there are multiple shortest paths between a pair, so betweenness centrality calculates the fraction of these shortest paths that go through the target person, and then adds all of the fractions from all possible pairs.
Triads are formed when two people you are connected to are also connected to each other. Triads show interconnected relationships within communities.
Who are the most peripheral people on these maps?
490 people are connected to only one person on these maps. Are these people unimportant? Do they really only work with one person in our network? The answer to these questions is a resounding no! Not everyone who is listed on these maps participated in the survey. These climate change professionals, who can be seen on the outer edges of the Fructerman Reingold map, are important enough to the network that our survey participants took the time to think of them and list their names as important contacts regarding climate change, weather, and the environment. They come from all different fields and from all over the Pacific and world. Almost certainly they have many connections with others in our network, and it is our hope that future studies can further capture their participation and connections.
A note about deceased, retired, and transferred colleagues.
The strength of our professional network in the Pacific Islands is impressive for a number of reasons, not the least of which is that it spans such a large area of the globe and includes many remote areas. However, because our network is so expansive, regular communications often rely on newsletters, listservs, and attendance at regional conferences. Throughout this project we have learned that news of our colleagues’ transfers, promotions, retirements, and even deaths, are not always communicated to the entire network (and had not always made it to us). This was an unexpected finding of the network analysis project. There is no mechanism for comprehensively communicating these events to colleagues in other countries or in other professions. In a network where months to years may pass between face-to-face meetings, it is not uncommon to celebrate a colleague’s promotion, or mourn their passing, well after the fact. To the best of our abilities, the information presented here is up to date as of the summer of 2013.
Confidentiality
If you zoom in on the high resolution images above, you may find that many people are listed by sector only, and not by their names. These people could not be contacted or did not give their permission for their names to appear, and so we are respecting their privacy.
Resilient and sustainable Pacific Island communities using climate information to manage risks and support practical decision-making about climate variability and change.
Full Network
The images on this page can be used to broadly understand the general structure of our professional network. For example, we see strong country clusters but we also see many strong connections between clusters. This will not be surprising to anyone in the Pacific—oftentimes our first conversation with a stranger is ‘who do we both know?’ to establish mutual friends and colleagues. In the first network map below, you can see broad trends of connection and centrality within the network. If you download the high resolution version of the second map below, you will be able to see individuals’ names and their connections to others. Exploring these maps, you can begin to look for connections that your friends and colleagues have. When considering future collaborations, you can explore who knows who within the region.
For example, to find potential collaborators by country:
Full Network Map 1: Participants by Country
Click here for High Resolution map image. Download the “Original (10000 x 10000)” size.
What is this map?
This map represents the full network findings of the Pacific RISA Network Analysis project. From December 2012 to April 2013, 331 climate change professionals from Hawaii and the U.S.-Affiliated Pacific Islands region, including our partners across the Pacific Islands region, the United States Mainland, and other international partners, completed a network survey. These 331 professionals noted connections with a total of 967 people that they communicate with regarding climate, weather, and environment issues. This map uses a layout algorithm called Fructerman Reingold, available in the Gephi Network Analysis software. This layout algorithm automatically brings people who are highly connected in the network to the center, and people with fewer connections fan out around the perimeter to create a circular web. This makes it easy to see how many people are centrally connected and how well they are connected to each other. The size of each circle indicates the number of connections each person has and people with the most connections to each other are pulled toward each other in the center of the map. The circles’ color indicates each person’s country or region.
What can we learn from it?
From this map, it is easy to see that the people who are very highly connected in the network have many connections to a) each other, and b) hundreds of other professionals in the region. Around the perimeter of the graph it is easy to see where central people from a country or region connect with others from their country or region. In the center of the graph, the apparent clusters of color (e.g., the clusters of light blue circles, orange circles, and dark purple circles) indicate that people in each country or region are strongly interconnected. However, the fact that the clusters are so close to each other, and overlap with each other, indicates that strong connections in our region transcend locational boundaries.
Full Network Map 2: Participants by Country
Click here for High Resolution map image. Download the “Original (10024 x 10024)” size.
What is this map?
This map is composed of the exact same people and connections as the map above. This map uses a layout algorithm called Force Atlas 2, also available in the Gephi Network Analysis software. This layout pulls clusters apart so that it is easier to see who is connected with whom in a cluster, and how well the clusters are connected with each other. Again, the circles are sized according to connectedness and colored according to country or region.
What can we learn from it?
What was hinted at in the Fructerman Reingold map (Full Network Map 1, see above) becomes very clear here—our professional network has strong country clusters. It is also quite apparent that these clusters are very highly interconnected. There is no single person or small group of people, for example, linking American Sāmoa (orange) with the rest of the region. Our network is highly connected across spatial boundaries—we are a series of islands connected in our professions.
Who are the most central people on these maps?
Centrality can be measured in different ways; each measure means something different. The following table lists the five “most central” people in the full network from each region, according to different measures of centrality. See below for definitions.
Degree centrality is the number of people you have listed connections to in this region. It is measured from 0 (no connections) to the network population minus one.
Eigenvector centrality looks at a person’s position within the network, basically measuring each person’s centrality according to the centrality of their connections. It is measured from 0 to 1 (highest eigenvector centrality in the network).
Closeness centrality is the inverse of farness, which is the sum of how many hops one must make to connect to all others in the network. Our network is pretty interconnected, so closeness centrality in this network ranges from 2.03 to 5.33 (and more than 2/3 of our network has a closeness centrality of 3.5 or lower).
Betweenness centrality considers how many shortest paths between pairs of people across the network pass through a given person. Oftentimes, there are multiple shortest paths between a pair, so betweenness centrality calculates the fraction of these shortest paths that go through the target person, and then adds all of the fractions from all possible pairs.
Triads are formed when two people you are connected to are also connected to each other. Triads show interconnected relationships within communities.
Who are the most peripheral people on these maps?
490 people are connected to only one person on these maps. Are these people unimportant? Do they really only work with one person in our network? The answer to these questions is a resounding no! Not everyone who is listed on these maps participated in the survey. These climate change professionals, who can be seen on the outer edges of the Fructerman Reingold map, are important enough to the network that our survey participants took the time to think of them and list their names as important contacts regarding climate change, weather, and the environment. They come from all different fields and from all over the Pacific and world. Almost certainly they have many connections with others in our network, and it is our hope that future studies can further capture their participation and connections.
A note about deceased, retired, and transferred colleagues.
The strength of our professional network in the Pacific Islands is impressive for a number of reasons, not the least of which is that it spans such a large area of the globe and includes many remote areas. However, because our network is so expansive, regular communications often rely on newsletters, listservs, and attendance at regional conferences. Throughout this project we have learned that news of our colleagues’ transfers, promotions, retirements, and even deaths, are not always communicated to the entire network (and had not always made it to us). This was an unexpected finding of the network analysis project. There is no mechanism for comprehensively communicating these events to colleagues in other countries or in other professions. In a network where months to years may pass between face-to-face meetings, it is not uncommon to celebrate a colleague’s promotion, or mourn their passing, well after the fact. To the best of our abilities, the information presented here is up to date as of the summer of 2013.
Confidentiality
If you zoom in on the high resolution images above, you may find that many people are listed by sector only, and not by their names. These people could not be contacted or did not give their permission for their names to appear, and so we are respecting their privacy.
Our Vision
Resilient and sustainable Pacific Island communities using climate information to manage risks and support practical decision-making about climate variability and change.
Tag Cloud
Follow Pacific RISA
Subscribe to our mailing list
Archives
Links