ICESat TM 04/21/2003 1

MIT Activities

• Two main areas of activity:– Validation of atmospheric delays being computed

for ICESat– Assessment of statistics of return signal strengths.

• Analysis of max received amplitude (maxRecAmp GLA05)

• Analysis of Area under received return (areaRecWF1 GLA05)

• Analysis of “smooth” topography slope. RMS about 1-sec linear trends in topography less than 20-cm (elev GLA06)

• Addition to the Atmospheric ATBD to evaluate large off nadir pointing (increased from 10 to 35 degrees)

ICESat TM 04/21/2003 2

Atmospheric Delay Validation

• Comparison of MIT developed software with the GLA06 dry tropospheric delay.

• MIT obtains the 3-D analysis fields from NCEP ftp area. We use the final assimilation model (GRIB format, extent .fnl; other files contain for example the aviation format but we do not use these.

• Comparison: Along the GLAS ground track (based on GLA06), compute dry delay at the GLA06 ellipsoidal height+geoid height (also from GLA06).

• Differences are not understood yet.

ICESat TM 04/21/2003 3

Comparison of MIT and GLA06 dry delay

ICESat TM 04/21/2003 4

Zoom of MSL interval

ICESat TM 04/21/2003 5

Differences between MIT and GLA06

ICESat TM 04/21/2003 6

Comparison of Differences

• No apparent height dependence of difference. Largest differences are at MSL

• MIT code checked against the surface pressure in the GRIB files (normal calculation downwards continues the 1000 mbar surface to avoid some ringing at NCEP surface pressure after rapid changes

• Most likely explanation is that GLA06 is generated from a different assimilation model.

• There seems to be some small interpolation problems in GLA06

ICESat TM 04/21/2003 7

Signal Strength Studies

• Three class of studies carried out (two use GLA05, one uses GLA06)– Peak received voltage. Large percentage are reaching limit of

1.5 volts, peaks in Gaussians have similar (and often larger values).

– Area under the received waveform. Values are reaching the threshold of I*2 storage of values. Also some negative values

– Three basic classes of large areas:• Strong sharp pulse• Long duration pulses with relatively low amplitude but integrate to

large area (clouds)• Not clear cases: Large area, by short pulse with moderate

amplitude

– Slopes of smooth terrain (defined by small RMS fit to linear trend over 1-sec).

ICESat TM 04/21/2003 8

Peak signal Voltage

ICESat TM 04/21/2003 9

Histogram of Max Received Signal

ICESat TM 04/21/2003 10

Waveform Area (mVolt-ns)

ICESat TM 04/21/2003 11

Histogram of values (log scale)

ICESat TM 04/21/2003 12

Signal Strength and surface shape

• The peak return voltage is clearly being clipped and the often even over clouds we are seeing very strong returns (although often with long return pulses)

• Integrated power under waveform is also large although the distribution in deceasing with increasing size

• No re-normalization for receiver gain yet. (Scaling of gain counts?)

• Now examine surface slope and RMS fit to 1-sec data. Looking for areas of specula reflections.

ICESat TM 04/21/2003 13

GLA06_030316 Ground Track

ICESat TM 04/21/2003 14

GLA06 elevation plot

ICESat TM 04/21/2003 15

GLA06 Slope and RMS Analysis

• Analysis looks at the slope and RMS of GLA06 elevation values by taking the data in 1-sec samples and fit a linear trend and computing the RMS about the trend.

• Most of the RMS fits < 20 cm are over the oceans and so far all the slopes have been small (<10 milli-deg). RMS can be as low as 25 mm over calm oceans.

• Slopes comparable to the ICESat off-nadir pointing angle and small RMS seen on ice sheets and some continental areas. If these surfaces are capable of specula reflections then slopes are a problem.

• Lots of interesting features in ocean returns

ICESat TM 04/21/2003 16

RMS and Slopes

ICESat TM 04/21/2003 17

Zoom ice area

ICESat TM 04/21/2003 18

High slope area of ice sheet

ICESat TM 04/21/2003 19

Summary

• Some work still to do on validation of tropospheric delay– Most likely different pressure analyses being used.

Differences may indicate overall quality of atmospheric delay

• Signal Strengths– 9-10% of pulse reach the 1.5Volt limit on return signal

strength– Issue with maximum possible signal at detectors:

• Gain changes not accounted for yet (need to know scaling): Concern if gain high and strong returns comes in

• Area under received pulse still being studied

• Many areas of 2-20cm elevation RMS; many in oceans where slopes seems to small (as we have seen) but some on ice sheets where slopes can be similar to current off-nadir pointing angle.

ICESat TM 04/21/2003 20

GLA05 GLA06 Data files

• Overall the quality of data and file contents is excellent.

• Alignment of GLA05 and GLA06 data files not clear at present. Day 2003 075: Three GLA05 data files that do no overlap with the one GLA06 file.

• Some anomalies in record entries and mismatches between entries of GLA05 and GLA05 (e.g, overflow of Area but Gaussian fits are infinite, number of peaks negative in GLA06 in many cases). These anomalies will be reported.