CERES/GEO and CERES-only(ERBE) LW Diurnal Averaging Comparison(1)
- CERES-only(ERBE) LW land temporal interpolation uses a half-sine and constant nighttime fit if the
daytime LW flux is greater than the nighttime flux, to fill in 24 hourly time steps during the day
- CERES-only(ERBE) LW ocean and snow temporal interpolation uses linear interpolation, to fill in 24 hourly
time steps during the day. It is assumed that there are no LW diurnal cycles over clear-sky ocean and snow surfaces
- CERES/GEO LW temporal interpolation uses all CERES and 8 3-hourly (GMT) GEO derived BB fluxes normalized at
coincident CERES/GEO measurements to maintain CERES instrument calibration. All remaining hourly time steps
are linearly interpolated between measurement times. For nonpolar regions there are usually
2 CERES and 8 GEO measurements, almost half the hourly time steps
- The CERES-only(ERBE) land LW half-sin fit models the daytime heating, which assumes symmetry about noon.
- Note, that for a given daytime measurement, both a previous and following nighttime observation must be present to apply
half-sine fit. If the conditions are not met linear interpolation between measurements are used,
causing erroneous flux estimates.
- It seems that the ERBE half-sine fit does not capture the daytime heating lag and nighttime cooling over land.
CERES/GEO and CERES-only(ERBE) LW Diurnal Averaging Comparison(2)
- CERES-only(ERBE) LW interpolation would show symmetry about noon or no difference between 16:30 and 7:30.
- CERES/GEO clearly show the land afternoon convection (blue regions), which are colder in the afternoon than
morning and daytime heating thermal lag (red regions), indicating land is warmer near sunset rather than sunrise.
- Also the Peruvian stratus regions are warmer in the afternoon indicating the clear-sky ocean is warmer
than the morning cloud tops
- PM-AM differences can be ~ 30 W/m2
CERES/GEO and CERES-only(ERBE) LW Diurnal Averaging Comparison(3)
- Difference of the monthly regional means of CERES/GEO and CERES-only temporal SW averaging
for Terra (10:30 AM sun-synchronous) December 2002
- Terra-CERES-only monthly means are similar to the CERES/GEO LW temporal averaging because the daytime and
nighttime LW biases cancel out.
- CERES/GEO LW interpolation averaging reveals 3-hourly or 45° longitude striping due to discritization
of the 3-hourly GEO measurements.
- Global bias is +0.5 W/m2.
CERES Data Product Information