Balancing horizontal accuracy and data collection efficiency with mapping-grade GPS receivers

Weusedmapping-gradeglobalpositioningsystem(GPS)receiverstoexaminepost-processedhorizontalmeasurement accuraciesduring nine separate visitstosix sitesthat eachcontained 10 test points withina second-growth Douglas-firforest.OurprimaryobjectivewastocompareGPSaccuraciesresultingfromtwodifferentreceiverconfigurations that vary in productivityand accuracy. A secondaryobjective wasto determine howaccuracies would change with recordinginterval (whether we averaged 1, 5, 10, 30 or 60 positions). We also examined which environment-andGPS-relatedfactorsmostinfluencedGPSaccuracy.Rootmeansquareerror(RMSE)atthe95percent confidencelevel(RMSE95)usingthe‘default’GPSsetting(positiondilutionofprecision(PDOP):6;SNR39;minimum elevation158)was6.0 m,whileRMSE95using ‘maximum’settings(PDOP: 20;SNR33;minimumelevation:58)was 8.0 macrossallsitesandrecordingintervals.Usingdefaultsettings,RMSE95decreasedfrom7.3 mwhenaveraging one position to 4.4 m when averaging 60. For maximum settings, RMSE95 decreased from 9.1 to 6.5 m. Our best models selected basal area, dilution of precision and weather, but not terrain. When using default settings the averagewait timewas 55 sec whencollecting one position and 236 sec whenaveraging60. Using maximum settings,averagewaittimewas8 secwhencollectingoneposition,and74 secwhenaveraging60positions.Theerrors that we report are acceptable for many forestry applications. These results should help users to balance productivity with accuracy when using mapping-grade GPS receivers in second-growth Douglas-fir forests.