Skip to Content
|30 Aug 2017 9:30 AM|
|Connect via zoom from any computer.|
The Massey Geoinformatics Collaboratory presents:
Is Remote Sensing Useful for Finding and Monitoring Urban Farms: Combining Data Mining and Geospatial Methods for Mapping Peri-Urban Agriculture in Detroit, Harare, Dakar, and Ho Chi Minh City
Presenters: M.E. Brown1, J.L. McCarty2,3
1.Department of Geographical Sciences, University of Maryland College Park, MD, firstname.lastname@example.org
2.Michigan Tech Research Institute, Ann Arbor, MI, email@example.com
3.Department of Geography, Miami University, Oxford, OH, firstname.lastname@example.org
Abstract: As the world becomes increasingly urbanized, the need for fresh fruits and vegetables in urban areas grows while the difficulty of timely provision of these perishable products also increases. Small-scale agriculture located in urban areas is a highly effective and profitable way to provide fresh produce to communities that are far from extensive commercial agricultural areas. Here we describe how remote sensing can be used with data mining approaches to monitor urban and peri-urban farms within cities in both developed and developing countries. Using very high resolution satellite imagery together with moderate and coarse resolution imagery and information from social media and the web, we analyze the usefulness of different methods to identify farms within urban boundaries in four countries. Data mining techniques were necessary to discover known urban farms within four cities, Detroit, Michigan, USA; Ho Chi Minh City, Vietnam; Harare, Zimbabwe; and Dakar, Senegal for use as training sites to compare the remote sensing monitoring and identification methods. Social media platforms included Twitter, Facebook, and Instagram, with generic and geographically targeted web scraping techniques to extract business names and locations from existing urban farm websites and online databases. A mixed-method approach was necessary in order to identify where urban farming is occurring and to monitor its change through time. Although remote sensing-based vegetation and water indices were useful, without ancillary data they are not effective at identifying the locations of urban farms. Remote sensing is a good way to monitor vegetation condition in locations where actively managed urban farms, but knowledge of locations of existing peri-urban agriculture is necessary.
Date and time: 9:30am on Wednesday 30th August NZ time.
The zoom link is: https://massey.zoom.us/j/842549186.
All are welcome, and please feel free to pass this invitation on to others who may be interested.
Page authorised by Manager, Events
Last updated on Tuesday 16 August 2016