Posts Tagged ‘Google Earth’

Temporal geographic information: a work in progress

Friday, March 23rd, 2012

The importance of time in Geography has become more relevant for me as I began working on my research project. It was helpful to read Langran and Chrisman’s (1988) article. In a way I was comforted to relate to some of the issues with regards to dealing with time, but at the same time felt discomfort that these issues are still around. We live in a digital geographic world, as sah mentioned in their post. Andrew stated PPGIS and HCI display the issues that arise when using Google applications. Currently dealing with the LBS and open-source world where everything is rapidly changing, new versions of software quickly replacing the old, past problems quickly become obsolete. However, I have learned from this article that time, like other fundamental concepts in Geography, is different. It is still a timely (no pun intended) issue. So how do we go about dealing with mapping time, along with theme and location (1)? Although still in its early stages, the Ushahidi platform may fulfill the requirement of being “a temporal database that makes the time dimension accessible to users” in  the example give by the Ghana Waters initiative (2).

The space-time composite section reminded me of the problem of overlaying a polygon layer created in ArcMap. For example, a geographer decided to represent suburbs of a city by digitally drawing polygons. If they want to display this as a choropleth map, displaying crimes throughout the city, they can do so. However, over time, the boundaries of suburbs may change, thus a new layer must be created to ensure timely accuracy of the theme and space that is represented. I believe the advantage of having Google Earth now, as opposed to 1988, is that we can integrate conventional software databases like ArcGIS with user-friendly, interactive virtual globes to try and solve time related problems. Altering between suburb overlay choropleths from one time period to another can be done by checking a box. Creating a time-lapse animation could be a possible solution to static images that “do not represent the events that change one state to the next” (8). It’s still a work in progress, however, the less constraints we have when dealing with more philosophical and abstract concepts such as time (and ofcourse, ontologies), the better.

-henry miller

Coarse grained data issues low resource settings

Friday, March 16th, 2012

Despite Goodchild et al.’s (1998) article’s technical components, the article did make me think of uncertainty regarding boundaries and course grained satellite imagery. Exploring low resource settings on Google Earth is one such example. Although an incomplete geolibrary, I consider Google Earth to be effective in its user friendly interface and features (layers and photographs), and of course, ubiquity. It’s a start. With this in mind, ‘flying’ over towns in Colombia on Google Earth, and the terrible, terrible satellite imagery that was available. (The low quality imagery remains unchanged since the last time I checked it half a year ago). One of the towns/districts is Puerto Gaitan. How do we account for the lack of resources given to collecting fine grained even medium grained visualizations?

According to Goodchild et al., alternative methods for displaying fuzzy regions must be applied where cartographic techniques are not enough. “A dashed region boundary would be easy to draw, but it would not communicate the amount of positional uncertainty or anything about the form of the z(X) surface” (208). What do we do then, when the data cannot even be analyzed because it is too coarse? For low resource settings, we are just going back to where we started. No financial incentives to improve data (from coarse to fine) = continuation of coarse grained data = poor visualization = cannot be utilized in studies = no advancements in research are made = back to the start, no financial incentives to improve the quality of data. How do we break this cycle?

-henry miller