1. We can also see i n Figwe 2 that the absolute RSS errors (the position errors of spacecraft 1 or spacecraft 2) grow rapidly between landmark observations and drop sharply a t the landmark observations. lhey reach their largest values a t the end of lone gaps in landmark data, before landmarks 3 and 4 and at the end of the rum, and take on their minimm values in the 120-160 minute time interval. niis behavior also holds for a l l the rms i n 3ibl.e 2, except that the curves are much smoother. and the magnitudes much larger, for rrns without intersatellite angle data. It is more difficult to make general statements about the relative RSS errors, since even the qualitative behavior vavies among the different rums. h e maxirnm relative RSS values in Table 2 are cmputed for times greater than 40 minutes, i n an attempt t o minimize the effects of transients, but are still not as reliable as the other maxima and minima i n the table.
2. h e geopositioning error curves i n Figure 3 show more complex behavior than Figwe 2. Some simplication r e s u l t s from r e s t r i c t i n g the discussion t o the maximim and minimm error cwves, which represent the major and minor semiaxis lengths of the geopositioning error ellipse. and ignoring the less interesting east/west and north/ south error curves. The maximum and minimum geopositioning errors are also the errors i n the intrack (along the spacecraft ground track) and crosstraok directions, respectively. h i s can be Seen from the relationship of the two sets of cutves in Figure 3, espeaially at 26, 75, 124, and 173 minutes, when the gramd track reaches maximun north or south latitude, and the east/west and 'd intrack directions coincide. The intrack and crosstrack geopositioning errors grow rapidly between landmark observations, drop sharply a t landmark observations, aod take on their minimum values i n the 120-160 minute int,erval, like the absolute RSS errw curve 9 i n Figure 2. lhe geopositioning error curves also exhibit jmps a t the day/night (DiN) and nightlday (Nil)) transitions when t h e Spacecraft ground track crosses the terminator; these jumps are due solely t o the different daytime and nighttime resolutions assumed for the earth Senmr. nie maximum intrack geopositioning errors occur a t the Same times as the maximum absolute RSS errors. but the maximum crosst r a c k geopositioning e r r o r s are found near 90 minutes. A l l these results generalize t o a l l the runs i n Table 2, except that the maximum crosstrack errors occur at closer t@70 minutes for the noncoplanar (Q, = 3.8 degree) rums with intersatellite angle data.