Summary of Data Analysis: Low-Band 1, Extended Ground Plane · 4.Receiver calibration S11 file:...
Transcript of Summary of Data Analysis: Low-Band 1, Extended Ground Plane · 4.Receiver calibration S11 file:...
LOCO EDGES REPORT #106
Summary of Data Analysis: Low-Band 1, Extended Ground Plane
Raul A. [email protected]
CASA, University of Colorado BoulderSESE, Arizona State University
August 31, 2017
Here we show results for the analysis of data from the Low-Band 1, Extended Ground Plane instrument.An equivalent report for the Low-Band 1, Original Ground Plane instrument, can be found here:
http://loco.lab.asu.edu/loco-memos/edges_reports/report105.pdf
Nominal choices and calibration settings:
1. Dates: 2016-258 to 2017-95
2. Sun cut: none
3. Moon cut: none
4. Receiver calibration S11 file: s11 calibration low band LNA25degC 2015-09-16-12-30-29 simulator2 long.txt
5. Receiver parameter polynomial terms: Nfit=6, Wfit=5
6. Antenna S11 file: S11 blade low band 2016 243.txt
7. Antenna S11 modeling: 9 polynomial terms after removal of delay
8. Balun loss correction: yes
9. Ground loss correction: yes, 0.5%
10. Beam correction: yes, using beam file newniv rotated to AZ = −7◦, and Haslam sky map scaled to76 MHz using β = −2.5.
1
The results are summarized in the following figures.
The first nine figures (after Figure 1 described below) provide a overview of the data.
• Figure 2: daily residuals to 7-term polynomial over 50-100 MHz, for all days considered. The GHArange is 0-24 hr.
• Figure 3: residuals for 4-hr GHA averages, to 7-term polynomial over 50-100 MHz.
• Figure 4: residuals for 6-hr GHA averages, to 6-term polynomial over 65-95 MHz. Averaging over 6hours produces estimates that are more stable as a function of GHA.
• Figure 5: residuals for 6-hr GHA averages, to 6-term polynomial plus 21-cm signature, over 65-95MHz. The signature center frequency was fixed at 78.5 MHz.
• Figure 6: amplitude of 21-cm signature as a function of GHA for 6-hour averages, in parallel to thetotal sky temperature. The signature center was fixed at 78.5 MHz.
• Figure 7: residuals and model for average over GHA 0-24 hr, to 1) 7-term polynomial over 50-100MHz; 2) 7-term polynomial over 53-97 MHz; 3) 7-term polynomial plus 21-cm signature over 53-97MHz.
• Figure 8: residuals and model for average over GHA 0-24 hr, to 1) 7-term polynomial over 50-100MHz; 2) 7-term polynomial over 60-100 MHz; 3) 7-term polynomial plus 21-cm signature over 60-100 MHz.
• Figure 9: residuals and model for average over GHA 0-24 hr, to 1) 6-term polynomial over 50-100MHz; 2) 6-term polynomial over 60-98 MHz; 3) 6-term polynomial plus 21-cm signature over 60-98MHz.
• Figure 10: residuals and model for average over GHA 6-18 hr, to 1) 6-term polynomial over 50-100MHz; 2) 6-term polynomial over 60-98 MHz; 3) 6-term polynomial plus 21-cm signature over 60-98MHz.
The next five figures show the results for the nominal data average, and compare them with those ob-tained for different data cuts and calibrations. All the cases correspond to data averages over GHA 6-18hr and 60-98 MHz, modeled with a 6-term polynomial plus a 21-cm signature. In these cases, the fre-quency and duration of the 21-cm signature, as well as its amplitude, are found by maximizing the ratioSNR = |a21/σ21|.• Figure 11: residuals and 21-cm signature for three different averages of consecutive subsets of data.
• Figure 12: residuals and 21-cm signature for four different averages: 1) daytime only (ELsun ≥ 0◦),2) nightime only (ELsun < 0◦), 3) Moon up only (ELmoon ≥ 0◦), 4) Moon down only (ELmoon < 0◦).
• Figure 13: residuals and 21-cm signature for two alternative receiver calibrations: 1) cal2015, Nfit=7,Wfit=7, and 2) cal2017, Nfit=7, Wfit=7.
• Figure 14: residuals and 21-cm signature for three alternative antenna S11: 1) 2017-93 (2015 switchparameters), 2) 2016-243 (2017 switch parameters), 3) 2017-93 (2017 switch parameters).
• Figure 15: residuals and 21-cm signature for two alternative corrections: 1) applying no ground losscorrection, and 2) applying no beam correction.
Finally, Table 1 compares the estimates for the different data cuts and calibration alternatives, and Figure1 shows the values.
2
Table 1: Signature amplitude estimates for all the cases considered. In all cases, the GHA range is 6-18 hr.The frequency range is 60-98 MHz and, in addition to the signature, the fit model includes six polynomialterms.
Case Description Amplitude estimate [mK]
0 Nominal 447± 161 1st data subset 428± 272 2st data subset 446± 233 3st data subset 515± 264 Daytime only 493± 345 Nighttime only 452± 176 Moon up only 482± 187 Moon down only 448± 238 recv cal2015, Nfit=7, Wfit=7 571± 219 recv cal2017, Nfit=7, Wfit=7 503± 1710 ant S11 2017-93, switch2015 455± 1611 ant S11 2016-243, switch2017 465± 1612 ant S11 2017-93, switch2017 475± 1713 No ground loss correction 451± 1614 No beam correction 452± 17
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14case number
0
100
200
300
400
500
600
a21 [
mK
]
Figure 1: Signature amplitude for the conditions and cases of Table 1.
3
50 60 70 80 90 100frequency [MHz]
258259260261262263264265266267268269270271273274275276277278279280281282283285287288289290291292293294295296297298302303304
day o
f year
2016 [
1 K
per
div
isio
n]
176 mK
168 mK
181 mK
193 mK
187 mK
177 mK
174 mK
166 mK
199 mK
182 mK
202 mK
175 mK
182 mK
277 mK
185 mK
169 mK
175 mK
171 mK
184 mK
184 mK
165 mK
180 mK
160 mK
271 mK
247 mK
199 mK
189 mK
172 mK
210 mK
191 mK
145 mK
158 mK
175 mK
245 mK
175 mK
176 mK
184 mK
149 mK
157 mK
172 mK
152 mK
50 60 70 80 90 100frequency [MHz]
314315316317318319320321322323324327328329332334336337338339340341342343348350351352354355357358360361362363364365366
12
169 mK
151 mK
162 mK
152 mK
162 mK
175 mK
204 mK
144 mK
177 mK
166 mK
171 mK
175 mK
165 mK
218 mK
160 mK
163 mK
243 mK
224 mK
174 mK
194 mK
261 mK
169 mK
175 mK
204 mK
166 mK
172 mK
168 mK
177 mK
196 mK
196 mK
187 mK
181 mK
151 mK
174 mK
193 mK
296 mK
180 mK
192 mK
167 mK
235 mK
185 mK
50 60 70 80 90 100frequency [MHz]
356789
11121316177778798082838485868788899394
170 mK
183 mK
163 mK
193 mK
151 mK
167 mK
158 mK
226 mK
186 mK
186 mK
137 mK
192 mK
166 mK
164 mK
199 mK
258 mK
209 mK
166 mK
178 mK
183 mK
179 mK
175 mK
214 mK
182 mK
185 mK
Figure 2: Daily residuals to 7-term polynomial over 50-100 MHz, for all days considered. The GHA range is 0-24hr.
50 60 70 80 90 100 110frequency [MHz]
20
16
12
8
4
0
GH
A [
0.5
K p
er
div
isio
n]
RMS: 135 mK
RMS: 83 mK
RMS: 57 mK
RMS: 63 mK
RMS: 102 mK
RMS: 144 mK
7 polynomial terms
Figure 3: Residuals for 4-hr GHA averages, to 7-term polynomial over 50-100 MHz.
4
50 60 70 80 90 100 110frequency [MHz]
18
12
6
0
GH
A [
0.5
K p
er
div
isio
n]
RMS: 46 mK
RMS: 34 mK
RMS: 34 mK
RMS: 41 mK
6 polynomial terms
Figure 4: Residuals for 6-hr GHA averages, to 6-term polynomial over 65-95 MHz.
50 60 70 80 90 100 110frequency [MHz]
18
12
6
0
GH
A [
0.5
K p
er
div
isio
n]
RMS: 37 mK
RMS: 22 mK
RMS: 12 mK
RMS: 26 mK
6 polynomial terms + signature
Figure 5: Residuals for 6-hr GHA averages, to 6-term polynomial plus 21-cm signature, over 65-95 MHz. Thesignature center frequency was fixed at 78.5 MHz.
5
0 6 12 18GHA [hr]
0
200
400
600
800
1000
1200
1400
a21
[m
K]
6 polynomial terms + signature
0
1000
2000
3000
4000
5000
T78.
5 [K
]
Figure 6: Amplitude of 21-cm signature as a function of GHA for 6-hour averages, in parallel to the total skytemperature at 78.5 MHz. The signature center frequency was fixed at 78.5 MHz.
50 60 70 80 90 100 110frequency [MHz]
[0.5
K p
er
div
isio
n]
RMS: 72 mK
RMS: 39 mK
RMS: 23 mK
τ= 7
a21 = 505 +/- 35 mK
νr = 78.2 MHz
∆ν= 19.2 MHz
7 polynomial terms, 0-24 GHA
Figure 7: Residuals for average over 0-24 GHA, to 1) 7-term polynomial over 50-100 MHz (blue); 2) 7-term polyno-mial over 53-97 MHz (green); 3) 7-term polynomial plus 21-cm signature over 53-97 MHz (red). Also shown (cyan)is the best fit 21-cm signature over 53-97 MHz.
6
50 60 70 80 90 100 110frequency [MHz]
[0.5
K p
er
div
isio
n]
RMS: 72 mK
RMS: 29 mK
RMS: 16 mK
τ= 7
a21 = 371 +/- 23 mK
νr = 78.2 MHz
∆ν= 18.0 MHz
7 polynomial terms, 0-24 GHA
Figure 8: Residuals for average over 0-24 GHA, to 1) 7-term polynomial over 50-100 MHz (blue); 2) 7-term poly-nomial over 60-100 MHz (green); 3) 7-term polynomial plus 21-cm signature over 60-100 MHz (red). Also shown(cyan) is the best fit 21-cm signature over 60-100 MHz.
50 60 70 80 90 100 110frequency [MHz]
[0.5
K p
er
div
isio
n]
RMS: 92 mK
RMS: 39 mK
RMS: 19 mK
τ= 7
a21 = 500 +/- 26 mK
νr = 78.0 MHz
∆ν= 19.8 MHz
6 polynomial terms, 0-24 GHA
Figure 9: Residuals for average over 0-24 GHA, to 1) 6-term polynomial over 50-100 MHz (blue); 2) 6-term polyno-mial over 60-98 MHz (green); 3) 6-term polynomial plus 21-cm signature over 60-98 MHz (red). Also shown (cyan)is the best fit 21-cm signature over 60-98 MHz.
7
50 60 70 80 90 100 110frequency [MHz]
[0.5
K p
er
div
isio
n]
RMS: 69 mK
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
6 polynomial terms, 6-18 GHA
Figure 10: Residuals for average over 6-18 GHA, to 1) 6-term polynomial over 50-100 MHz (blue); 2) 6-termpolynomial over 60-98 MHz (green); 3) 6-term polynomial plus 21-cm signature over 60-98 MHz (red). Also shown(cyan) is the best fit 21-cm signature over 60-98 MHz. The signature estimate shown here over 6-18 GHA, is close tothe estimate over 0-24 GHA, shown in the previous figure.
8
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Total Average
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 34 mK
RMS: 18 mK
τ= 7
a21 = 428 +/- 27 mK
νr = 78.2 MHz
∆ν= 19.8 MHz
Subset 1 of 3
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 34 mK
RMS: 17 mK
τ= 7
a21 = 446 +/- 23 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Subset 2 of 3
60 70 80 90 100 110frequency [MHz]
[0.2
K p
er
div
isio
n]
RMS: 37 mK
RMS: 17 mK
τ= 7
a21 = 515 +/- 26 mK
νr = 77.8 MHz
∆ν= 18.8 MHz
Subset 3 of 3
Figure 11: Residuals and 21-cm signature for three different averages of consecutive subsets of data. Each subsetcontains one third of the total data. Top panel is the nominal case.
9
60 70 80 90 100 110[0
.2 K
per
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Total Average
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 40 mK
RMS: 24 mK
τ= 7
a21 = 493 +/- 34 mK
νr = 77.8 MHz
∆ν= 19.6 MHz
Daytime Only
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 13 mK
τ= 7
a21 = 452 +/- 17 mK
νr = 78.0 MHz
∆ν= 19.0 MHz
Nighttime Only
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 35 mK
RMS: 14 mK
τ= 7
a21 = 482 +/- 18 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Moon Up Only
60 70 80 90 100 110frequency [MHz]
[0.2
K p
er
div
isio
n]
RMS: 33 mK
RMS: 17 mK
τ= 7
a21 = 448 +/- 23 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Moon Down Only
Figure 12: Residuals and 21-cm signature for four different averages: 1) daytime only (ELsun ≥ 0◦), 2) nightimeonly (ELsun < 0◦), 3) Moon up only (ELmoon ≥ 0◦), 4) Moon down only (ELmoon < 0◦). Top panel is the nominalcase.
10
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Nominal
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 42 mK
RMS: 17 mK
τ= 7
a21 = 571 +/- 21 mK
νr = 77.8 MHz
∆ν= 19.8 MHz
Calibration 2015, Nfit7 - Wfit7
60 70 80 90 100 110frequency [MHz]
[0.2
K p
er
div
isio
n]
RMS: 36 mK
RMS: 12 mK
τ= 7
a21 = 503 +/- 17 mK
νr = 77.8 MHz
∆ν= 19.6 MHz
Calibration 2017, Nfit7 - Wfit7
Figure 13: Residuals and 21-cm signature for two alternative receiver calibrations: 1) cal2015, Nfit=7, Wfit=7, and2) cal2017, Nfit=7, Wfit=7. Top panel is the nominal case.
11
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Nominal
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 11 mK
τ= 7
a21 = 455 +/- 16 mK
νr = 77.8 MHz
∆ν= 19.2 MHz
Antenna S11 2017-93
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 33 mK
RMS: 13 mK
τ= 7
a21 = 465 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.6 MHz
Antenna S11 2016-243, Switch2017
60 70 80 90 100 110frequency [MHz]
[0.2
K p
er
div
isio
n]
RMS: 34 mK
RMS: 13 mK
τ= 7
a21 = 475 +/- 17 mK
νr = 77.8 MHz
∆ν= 19.6 MHz
Antenna S11 2017-93, Switch2017
Figure 14: Residuals and 21-cm signature for three alternative antenna S11: 1) 2017-93 (2015 switch parameters),2) 2016-243 (2017 switch parameters), 3) 2017-93 (2017 switch parameters). Top panel is the nominal case.
12
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 447 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
Nominal
60 70 80 90 100 110
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 12 mK
τ= 7
a21 = 451 +/- 16 mK
νr = 78.0 MHz
∆ν= 19.4 MHz
No Ground Loss Correction
60 70 80 90 100 110frequency [MHz]
[0.2
K p
er
div
isio
n]
RMS: 32 mK
RMS: 13 mK
τ= 7
a21 = 452 +/- 17 mK
νr = 78.2 MHz
∆ν= 19.4 MHz
No Beam Correction
Figure 15: Residuals and 21-cm signature for two alternative corrections: 1) applying no ground loss correction, and2) applying no beam correction. Top panel is the nominal case.
13