About This Project

Understanding the environmental and human dimensions of the Amazon is essential to maintaining a balance between the need for resources and the preservation of the natural environment. In order to better understand these interactions, remote sensing and social media data will be integrated to help map the socioeconomic and physical environments. These maps will then be compared over time to determine if social media data can be used when examining land-use dynamics.

Ask the Scientists

What is the context of this research?

While many projects focus on the large scale environmental changes occurring in the Amazon, research using fine scale social and economic data is harder to find. Geospatial big data such as geocoded tweets from Twitter are quickly becoming an alternative dataset for spatial research. Such data allows us to uncover the spatiotemporal patterns of human activities by associating the semantic content of tweets with an activity. Functional boundaries can be inferred with this information in a process known as social sensing. This study will utilize social sensing via geocoded tweets and remotely sensed environmental data to map human activities in the Amazon in order to gain a better understanding of the land-use dynamics in the region.

What is the significance of this project?

With the ever-rising popularity of social media sites such as Twitter even people in the most remote places around the world are able to participate thus providing a new, more robust database of social information. Scientists can now utilize this socially sensed data that provides volunteered geographic information to analyze human behaviors and draw conclusions regarding the socioeconomic environment in these remote localities. Collecting and analyzing this type of information and combining it with environmental data is vitally important especially in places like the Brazilian Amazon where land use and land cover types are rapidly changing. Accurately mapping social and environmental data together could help governments make more informed land use planning decisions.

What are the goals of the project?

Using the Twitter API, we plan on collecting a data set of geo-located tweets within the Brazilian Amazon for one year. This data will be analyzed for semantic content and placed into categories based on human activities (e.g. transportation, recreation, education, etc.). A spatiotemporal map of the activity sectors with spatial statistical analysis results (hotspots/coldspots and density) over the study area will be generated. These land use sectors will be compared with environmental land use data obtained and classified using the latest remotely sensed satellite data (MODIS). The comparison of these maps will help us determine if Twitter can be used as a metric to better understand the land-use dynamics of the region and beyond.