14 



FISH AND WILDLIFE TECHNICAL REPORT 30 



Table 8. Precision of locations (in meters) obtained from transmitters at known locations. Fixed platform transmitter 

 terminals (PTT's) had long antennas and were not miniaturized. Only locations coded with quality indices 2 and 3 

 were obtained from fixed transmitters. Telonics PTT's were the type used on caribou (Rangifer tarandusj. Location 

 quality indices were not used by Argos before April 1987. 



deviations for locations coded as location quality (LQ) 3 at 

 Inuvik were 180 and 178 m (for comparison, the criterion 

 set by Argos for LQ3 is 150 m; Le Traon 1987). The 

 overall mean error from the Inuvik fixed PTT was 356 m, 

 the median was 271 m, and 90% of all locations were 

 within 713 m of the true location. 



Identical model PTT's at Inuvik and Nome were both 

 near sea level, but the PTT at Nome was exposed to greater 

 temperature fluctuations. Standard deviations from the 

 Nome PTT were somewhat greater than those from the 

 Inuvik PTT, varying from 10% greater for longitudinal 

 deviations from LQ3 locations, to 98% greater for longi- 

 tudinal deviations from LQ2 locations (Table 8), suggest- 

 ing that exposure to varying temperatures may have re- 

 duced precision. 



Mean error was smallest when satellite orbits had max- 

 imum elevations between 40 and 50 (Fig. 8). Greatest 

 error resulted from satellite orbits between 70 and 80; 

 however, these constituted a small (0.03) proportion of the 

 total overpasses. The most common overpasses had max- 

 imum elevations of 10-20. Theoretically, mean error 

 could be reduced by restricting locations to those resulting 

 from the best overpasses, but only with considerable loss 

 of data (Fig. 9). For example, using only overpasses >20 



and <70 improved mean precision by 6% but at a cost of 

 46% of the data. 



Location estimates from PTT's of the type used on 

 animals were less precise than were locations from fixed 

 PTT's having large, external antennas (Table 8). Standard 

 deviations from third-generation PTT's were 1.46-2.48 

 times those of the fixed transmitters. Standard deviations 

 from second-generation PTT's averaged 1.86-2.11 times 

 those of third-generation PTT's. 



Individual PTT's of the same type exhibited generally 

 similar precision. Four second-generation collars trans- 

 mitting during the same periods from a fixed location did 

 not differ in mean location error (Kruskal-Wallis H = 1 .43, 

 df = 3, n = 60), nor did four third-generation collars at the 

 same location (H = 1.92, df = 3, n = 1 13). Overall sim- 

 ilarity in performance did not mean that each overpass 

 resulted in identical performance from each PTT. Rather, 

 considerable variation was evident on an individual over- 

 pass basis (Fig. 10). Fancy etal. (1988) also presented data 

 that showed how location estimates obtained from indi- 

 vidual PTT's could vary considerably among a few 

 overpasses. 



Deployment on animals reduced precision from that 

 exhibited by PTT's while being tested at fixed locations 



