USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Improving Prairie Pond Counts with Aerial Video
and Global Positioning Systems

Results: Relationships of Pond Counts on Air-Ground Segments


The number of ponds counted on the video was positively correlated (r = 0.95, t = 9.70, P < 0.001) with the number of ponds (including Type I wetlands) counted by the ground crew on both sides of the 12 air-ground segments (Fig. 2). The slope of the regression of video counts on ground counts was different from 1 (t = 2.32, P = 0.043), and the intercept was not different from 0 (t = 1.77, P = 0.107). Video interpretation underestimated the number of ponds on the 2 air-ground segments with the largest pond counts. Omission error was greater than commission error at the 2 air-ground segments.

gif -- Graph of Video vs. Ground BGS Pond Counts

Figure 2. Number of ponds from video interpretation regressed on number of ponds from ground Aerial Waterfowl Breeding Ground Population and Habitat Surveys for 12 air-ground segments in southern Saskatchewan, May 1992. Pond counts are from both sides of air-ground segment and include Type I wetlands. Dashed line is for a relationship with a slope = 1 and an intercept = 0.

The number of ponds counted on the video was positively correlated (r = 0.98, t = 14.89, P < 0.001) with the number of ponds (excluding Type I wetlands) counted by the ground BGS on 1 side of the air-ground segment (Fig. 3a). The slope of the regression of video counts on ground counts was different from 1 (t = 2.92, P = 0.0153), and the intercept was different from 0 (t = 3.10, P = 0.0112). Video interpretation tended to overestimate the number of ponds at air-ground segments with intermediate numbers of ponds and underestimate the number of ponds at air-ground segments with large pond counts.

gif -- Graph of Video vs.Ground BGS Pond Counts

Figure 3a. Relationships of pond counts from video and ground Aerial Waterfowl Breeding Ground Population and Habitat Surveys (BGS). Pond counts are from the one side of 12 air-ground segments counted in the aerial BGS in southern Saskatchewan, May 1992. Type I wetlands identified by the ground BGS are excluded from the ground and video counts. Dashed line is for a relationship with a slope = 1 and an intercept = 0.

The number of ponds counted on aerial BGS was positively correlated (r = 0.97, t = 13.42, P < 0.001) with number of ponds (excluding Type I wetlands) counted by ground BGS on 1 side of the 12 air-ground segments (Fig 3b). The slope of the regression of ocular aerial counts on ground counts was different from 1 (t = 3.67, P = 0.004), and the intercept was not different from 0 (t = 1.04, P = 0.322). The ocular aerial count tended to underestimate the number of ponds.

gif -- Graph of Aerial vs. Ground BGS Pond Counts

Figure 3b. Relationships of pond counts from aerial and ground Aerial Waterfowl Breeding Ground Polockquotepulation and Habitat Surveys (BGS). Pond counts are from the one side of 12 air-ground segments counted in the aerial BGS in southern Saskatchewan, May 1992. Type I wetlands identified by the ground BGS are excluded from the ground and video counts. Dashed line is for a relationship with a slope = 1 and an intercept = 0.

The number of ponds counted on aerial BGS was positively correlated (r = 0.93, t = 8.73, P < 0.001) with number of ponds (excluding Type 1 wetlands) counted on video (Fig. 3c). The slope of the regression of ocular aerial counts on video counts was not different from 1 (t = 1.11, P = 0.294), and the intercept was not different from 0 (t = -0.59, P = 0.569). Although these statistics suggest there was no consistent tendency for aerial BGS to over- or underestimate the number of ponds counted in the video interpretation, the scatter plot shows the ocular aerial pond count was less than the video count for 8 of 12 air-ground segments.

gif -- Graph of Aerial BGS vs. Video Pond Counts

Figure 3c. Relationships of pond counts from video and aerial survey of the Aerial Waterfowl Breeding Ground Population and Habitat Surveys (BGS). Pond counts are from the one side of 12 air-ground segments counted in the aerial BGS in southern Saskatchewan, May 1992. Type I wetlands identified by the ground BGS are excluded from the ground and video counts. Dashed line is for a relationship with a slope = 1 and an intercept = 0.


Previous Section -- Results: Errors in Video Interpretation of Ponds
Return to Contents
Next Section -- Results: Relationships of Pond Counts on Transect Segments

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo USA.gov logo U.S. Department of the Interior | U.S. Geological Survey
URL: http://www.npwrc.usgs.gov/resource/methods/pndcount/agseg.htm
Page Contact Information: Webmaster
Page Last Modified: Saturday, 02-Feb-2013 06:02:33 EST
Sioux Falls, SD [sdww54]