The NEPs measured in orbit are higher in general than measured on the
ground. (See Selected Notes for LWS for the reasons).
The following almost unreadable plot shows how they have changed at each
wavelength. The plots are for sources with fluxes of 10^-14, 10^-15, 10^-16
W/m^2 in a resolution element, from top to bottom. The y-axis value plotted
is the ratio of the in-orbit estimate of the S/N in 1 sec vs. the ground
value (as per the time estimator and PGA). The x-axis is labeled 60, 80,
100, 120, 140, 160, 180 um. The y-axis is a log scale with labels 1.00
(top), 0.10 (middle), and 0.01 (bottom). One can see (barely) that for
the lowest box, 10^-16 W/m^2 in a resolution element, the S/N has dropped
by factors of 2 to 8. In the top box the S/N has dropped slightly more
at long wavelengths. The sensitivity is a function of the source strength.
For the brightest sources, with fluxes greater than 10^-14 W/m^2 in a resolution
element, the change in NEP begins to become dominated by maximum S/N obtainable
in an integration ramp. The current (March 1996) estimate is that a S/N
of 1/400 is about the maximum that can be obtained from one (1/2 second)
ramp. Previously, this was set to be as high as S/N=1000 in the time estimator.
Users should run the current time estimator to determine estimated sensitivities
that result from this limit.