Harvard Forest >

Harvard Forest Symposium Abstract 2013

• Title: Consolidation of Harvard Forest Remote Sensing resources
• Primary Author: Paul Siqueira (University of Massachusetts - Amherst )
• Additional Authors: Brian Hall (Harvard Forest)
• Abstract:

  Serving as a component of NSF’s Long Term Ecological Research Network since 1988, the Harvard Forest has a long history of serving the ecological and natural science communities by being a hub of experimental and observational research. Because of its involvement with the LTER program, a vast expansion in remote sensing technology and the need for environmental monitoring on regional to continental scales, a significant amount of remote sensing data have been collected over the Harvard Forest during this time. The mapping resources that these data represent complements those that have been collected by LTER researchers and stored in the Harvard Forest data archive. The more dispersed and sensor-specific nature of remote sensing data however have led to a variety of investigations and data sources that have not been as well centralized as some of the other historic datasets being maintained through the facilities’ archival resources. Hence access to, and knowledge of these resources has been a challenging and inefficient task that limits the larger goal of the Harvard Forest LTER as being an open-access facility.
  
  
  
  
  
  For this reason, we have been working on consolidating remote sensing resources (and results) that have been collected over the years at the Harvard Forest and to provide a mechanism for the interested user to access those data. The types of remote sensing data that are intended to be made more easily available are
  
  
  
  
  
  *.) observational lidar datasets that determine canopy height and below-canopy ground surface topography,
  
  
  
  
  
  *.) hyperspectral data that are responsive to reflectance bands throughout the optical an infrared spectrum from 400 to 2500 nm.
  
  
  
  
  
  *.) microwave observations from synthetic aperture radar that provide high-resolution (~10m) reflectance imagery over the wider region
  
  
  
  
  
  *.) landsat multispectral bands ranging from 450 nm to 12500 nm.
  
  
  
  
  
  While some of these resources are readily available through national facilities like the USGS Landsat viewer (landsatlook.usgs.gov), other datasets collected by individual research organizations are more difficult to locate and access.
  
  
  
  
  
  The work that has been ongoing as part of this research work is to provide the foundation for collocating some or all of these data sets (or links to these data sets) onto a central location at the Harvard Forest, so that researchers can assess what types of remote sensing data are available and find ways to access them. The intention for the early stages of this project is to create enough of a resource, such that future additions to the resource will not require a large time investment by the individual researcher and that through its ease of access, will encourage others to augment the database with their own data and to use it for creating new and unanticipated applications.
  
  
  
  
  
  To achieve these goals, a prototype geographic information service has been created by ArcGIS Server that can be used to visualize available datasets. This same service is also capable of providing what are known as Web Mapping, Web Coverage and Web Feature Services (WMS, WCS and WFS, respectively) which can be automatically queried by software for the purposes of creating user-specified remote sensing data subsets that are constrained by sensor type, geographic or temporal extent. In short, these resources can be tailored by the user for the users needs, by the server, without requiring intimate knowledge of the mapping and geographic transformations that are necessary to create a useable resource.
  
  
  
  
  
  Below are shown three figures. Their captions are as follows:
  
  
  Figure 1. Demonstration of the Harvard Forest remote sensing database interface which can be accessed at http://107.20.174.13//HarvardForestRemoteSensingWebApplication/Default.aspx. Shown on this top layer is a map of canopy height, at a 1m spatial resolution, as observed by NASA Goddard's GLiHT instrument in the summer of 2012.
  
  
  
  
  
  Figure 2. A zoomed-in version (green rectangles in this figure and in Figure 1) of the GLiHT data focused on Shaler Hall. Geographically explicit data can be exported through html protocols known as WMS, WCS and WFS, so that the end-user can access the data for further processing.
  
  
  
  
  
  Figure 3. A block diagram of the Harvard Forest remote sensing resource data access and processing interface.
  
  
  

• Research Category: Regional Studies


• Figures:
Figure 1.jpg
Figure 2.jpg
Figure 3.jpg