Download - A Spreadsheet tool for the visualization of long term calibration … · 2014. 10. 27. · 6.- If it ain't broke, don't fix it! 7.- I am lazy, very lazy! The full calibration history

Transcript
Page 1: A Spreadsheet tool for the visualization of long term calibration … · 2014. 10. 27. · 6.- If it ain't broke, don't fix it! 7.- I am lazy, very lazy! The full calibration history

Monthly Mean Target Return Rate (PMT)

0

5

10

15

20

25

30

Au

g/2

01

1

Oc

t/2

011

Dec

/201

1

Fe

b/2

012

Ap

r/2

01

2

Ju

n/2

012

Au

g/2

01

2

Oc

t/2

012

Dec

/201

2

Fe

b/2

013

Ap

r/2

01

3

Ju

n/2

013

Au

g/2

01

3

Oc

t/2

013

Dec

/201

3

Fe

b/2

014

Ap

r/2

01

4

Ju

n/2

014

Au

g/2

01

4

Oc

t/2

014

Dec

/201

4

Fe

b/2

015

Ap

r/2

01

5

Rrate

Trend

A Spreadsheet Tool for the Visualization of Long Term Calibration Series ParametersA Spreadsheet Tool for the Visualization of Long Term Calibration Series ParametersJorge R. del Pino

Institute of Astronomy, University of Latvia Station 1884, [email protected]

Jorge R. del PinoInstitute of Astronomy, University of Latvia

Station 1884, [email protected]

19th International Laser Ranging Workshop“Celebrating 50 Years of SLR: Remembering the Past and Planning for the Future”

Annapolis, MD USA October 27-31/2014

Introduction

On creating the 1st generation DOS-based 10 Hz software for theSLR 7841 Potsdam, the goal was to extract and record themaximum possible information from all the observationalparameters.

It was decided to calculate and record, following the ILRS standards:• all the moments from the calibration set.• the derived statistical parameters.• the number of laser shots used, accepted for and filtered.• the calibration epoch.

The information is added in a single formatted line per calibrationin an detector dependent yearly file.This file is one the sources for the generation of the XML formatraw data input file for pass analysis.The same calibration information output format has been kept onthe kHz Linux-based software created by Spacetech.(http://www.spacetech-i.com/) for the SLR 7841, Potsdam.

Main Characteristics• A suite of Excel 97® spreadsheets with automatic links amongthem.• The inputs are Excel-compatible ASCII files generated by thepass filtering and target calibration programs.• Data transfer to the Excel® spreadsheets is by cut-and-paste(macros willbe added soon).• All graphical outputs are dynamical.

Statistic informationfor each parameter

Automatic alternatingcolor background forfaster month location

Date of 1st and last targetNumber of target calibrationsLaser shots (millions) on Dec 31stTarget calibrations on Dec 31st

The main page, the calibration data is pasted in columns C-N

Columns B, C & D are shown/hidden using Ctrl-o/Ctrl-cmacros for cut-and-pasting the data

3 groups of Excel® spreadsheets

Pass counting spreadsheets.

• Monthly and yearly pass results per type of satellite including: • Pass information for individual satellites. • Satellite categories and subcategories: Lageos, HEO, LEO, LEO geodetic, Tandem pairs, etc (All GNSS of a Network counted as one) • Days observed. • Passes/day. • Monthly and yearly mean RMS per satellite. • ILRS tracking goals. • End-of-year prognosis of number of passes and days of observation.

Daily observations spreadsheets. • For individual satellites, groups (LEO, HEO, Lageos) and true tandem passes. • Prognosis of when the ILRS goals can be/were reached for the LEO, HEO, Lageos categories.

Calibration results (by detector) spreadsheets.• Yearly and long term series of all statistical parameters.• Parameters Histograms.• Housekeeping statistics: • Results by day of week, month, year. • Number of days with N calibration. • Number and rates calibration/passes, monthly and yearly results.

The multiyear long term Excel® spreadsheets

A similar Excel can be generated with all the calibrations for agiven SLR configuration, for example the Potsdam 2004-201110 Hz version with 14000+ calibrations or the Riga 2001-2014recently generated with 10000+.

A global qualitative operational stability indexIf the station operates at single photon level, it is possible tocalculate and plot the calibration monthly mean return rate.This plot serves as a visual indicator of the global operationalstability of the system.Any change on the laser energy level, filter transmission, systemoptical alignment etc. will affect this mean return rate index.

If any Station or Network is interested on using or adapt these spreadsheets, contact me at: [email protected]

(Poster Abstract number 3111)

Examples of Housekeeping information

Why Excel 97®?

1.- I have a old registered copy of Office 97®.2.- The first spreadsheets versions were created in 2001.3.- This group of spreadsheets are compatible with any newer Office Version.4.- Using any newer Office version means learning a new layout for the commands.5.- This newer Office version will not improve the Excels functionality.6.- If it ain't broke, don't fix it!7.- I am lazy, veryvery lazy!

The full calibration history for Potsdam@10 Hz configuration.The long term stability is clearly visible.The 2008 jump is due to a PMT voltage change.

By filtering again the Riga raw calibration readout values, we could identifythe hardware changes and confirm them with the station journals.We could zeroed into a small time frame for searching the information.

1.- End of external target, a long multimode optical fiber as target.2.- Testing 2 exemplars of the Riga event timer.3.- Using a shorter length multimode optical fiber.4.- New shorter singlemode optical fiber as target.

Riga Raw Calibration Values 2001-2014

50000

150000

250000

350000

450000

550000

650000

750000

2000

-no

v-26

2001

-may

-13

2001

-oct

-28

2002

-ab

r-14

2002

-sep

-29

2003

-mar

-16

2003

-ag

o-3

1

2004

-feb

-15

2004

-ag

o-0

1

2005

-en

e-16

2005

-ju

l-03

2005

-dic

-18

2006

-ju

n-0

4

2006

-no

v-19

2007

-may

-06

2007

-oct

-21

2008

-ab

r-06

2008

-sep

-21

2009

-mar

-08

2009

-ag

o-2

3

2010

-feb

-07

2010

-ju

l-25

2011

-en

e-09

2011

-ju

n-2

6

2011

-dic

-11

2012

-may

-27

2012

-no

v-11

2013

-ab

r-28

2013

-oct

-13

2014

-mar

-30

2014

-sep

-14

pS

1

2

3

4

For what it is useful?

1.- Fast visualization of trends, jumps and data fluctuations.2.- Overview of the mean, median, σ, minimum and maximum values for each parameter.3.- Housekeeping information in numerical and graphical forms.4.- A set of standardized graphs, useful for reports, presentations and articles.5.- End-of-year prognosis on values.

Example of the Spacetech Linux-based calibration program output.The data framed in yellow is imported into the main Excel page

56658 1 64156 149094 11839 10616 90.383 0.049 0.040 0.08911 2.57 0.00118 0 1700 det1 las1 tim156658 1 67927 181155 11885 10493 90.394 0.047 0.038 0.10465 2.55 0.00091 0 1700 det1 las1 tim1

Uses of multiyear series information

Thanks to my colleagues in Potsdam & Riga, for their support and suggestions as users!

The auxiliary global tabulation Excel® spreadsheet

By automatically linking the different Excel® spreadsheets,the monthly pass and calibration information can be resumedin a tabular form.

The number of calibration days per month is also tabulated.This annual number of calibration days is close to, but notequal, to the total amount of clear days observed.

This poster is supported by the EU FP7 GRANT REGPOT-CT-2011-285912-FOTONIKA

Riga 10 Hz calibration 2003-2011

85.000

87.000

89.000

91.000

93.000

95.000

97.000

99.000

101.000

103.000

105.000

200

3-ag

o-2

3

2003

-dic

-13

200

4-ab

r-0

3

200

4-j

ul-

24

200

4-n

ov-

13

2005

-ma

r-05

2005

-ju

n-2

5

2005

-oct

-15

200

6-fe

b-0

4

2006

-ma

y-27

200

6-se

p-1

6

200

7-e

ne-

06

200

7-ab

r-2

8

200

7-ag

o-1

8

2007

-dic

-08

2008

-ma

r-29

200

8-j

ul-

19

200

8-n

ov-

08

200

9-fe

b-2

8

2009

-ju

n-2

0

2009

-oct

-10

201

0-e

ne-

30

2010

-ma

y-22

201

0-se

p-1

1

201

1-e

ne-

01

201

1-ab

r-2

3

201

1-ag

o-1

3

2011

-dic

-03

Potsdam 2014: Calibration Temporal Series

January-August trenddue to non-nominaloperation of the laseramplifier.

August: Full laserbreakdown

September: laserrestored to nominaloperation

Potsdam Target Calibration 2014

90.340

90.350

90.360

90.370

90.380

90.390

90.400

90.410

90.420

90.430

90.440

2013

-dic

-27

2014

-en

e-24

2014

-feb

-21

2014

-mar

-21

2014

-ab

r-18

2014

-may

-16

2014

-ju

n-1

3

2014

-ju

l-11

2014

-ag

o-0

8

2014

-sep

-05

2014

-oct

-03

2014

-oct

-31

2014

-no

v-28

2014

-dic

-26

Examples of Temporal Series, Potsdam 2014

Return Rate: % ofaccepted echoes duringcalibration.Acceptance rate: % ofechoes remaining afterfiltering the accepted ones.

0

10

20

30

40

50

60

70

80

90

100

110

27

-dic

-13

24

-en

e-1

4

21

-fe

b-1

4

21

-ma

r-1

4

18

-ab

r-1

4

16

-ma

y-1

4

13

-ju

n-1

4

11

-ju

l-1

4

08

-ag

o-1

4

05

-se

p-1

4

03

-oc

t-1

4

31

-oc

t-1

4

28

-no

v-1

4

26

-dic

-14

Acceptance Rate Return RateBias

-0.001

0.000

0.001

0.001

0.002

0.002

0.003

0.003

2013

-dic

-27

2014

-en

e-24

2014

-feb

-21

2014

-mar

-21

2014

-ab

r-18

2014

-may

-16

2014

-ju

n-1

3

2014

-ju

l-11

2014

-ag

o-0

8

2014

-sep

-05

2014

-oct

-03

2014

-oct

-31

2014

-no

v-28

2014

-dic

-26

Skewness

-0.05

0.00

0.05

0.10

0.15

0.20

0.25

2013

-dic

-27

2014

-en

e-24

2014

-feb

-21

2014

-mar

-21

2014

-ab

r-18

2014

-may

-16

2014

-ju

n-1

3

2014

-ju

l-11

2014

-ag

o-0

8

2014

-sep

-05

2014

-oct

-03

2014

-oct

-31

2014

-no

v-28

2014

-dic

-26

Kurtosis

2.48

2.50

2.52

2.54

2.56

2.58

2.60

2.62

2.64

2.66

2.68

201

4-e

ne

-06

201

4-f

eb

-03

201

4-m

ar-

03

201

4-m

ar-

31

201

4-a

br-

28

201

4-m

ay

-26

201

4-j

un

-23

20

14

-ju

l-2

1

201

4-a

go

-18

201

4-s

ep

-15

201

4-o

ct-

13

201

4-n

ov

-10

20

14

-dic

-08

Examples of Histograms

As the plots, they areautomaticallyredrawn every timenew data is added.

0

50

100

150

200

250

300

350

400

450

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95

Full Return Rate Acceptance Rate

%

Mean Deviation

0

100

200

300

400

500

600

0.01

0

0.01

2

0.01

4

0.01

6

0.01

8

0.02

0

0.02

2

0.02

4

0.02

6

0.02

8

0.03

0

0.03

2

0.03

4

0.03

6

0.03

8

0.04

0

0.04

2

In Riga there are data dips insummer (nights too short)and in winter (long nights/extensive cloud cover)Tracking only by night.

This is useful to schedulemaintenance stops orholidays while reducing itsimpact on the amount ofdata

Potsdam’s classicalweekend effect,fewer daylightobservationsSaturday/Sunday

The most common numberof calibrations for Potsdamin a 24 h period is ~5.5

WeekDay

2266 2271 2280

19831911

10681181

0

500

1000

1500

2000

2500

Mon Tue Wed Thu Fri Sat Sun

Number of Days with a value of N Targets/Day

0

10

20

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Daily Targets 2014

0

5

10

15

20

25

30

2014

-ene-

01

2014

-ene-

29

2014

-feb

-26

2014

-mar

-26

2014

-abr-

23

2014

-may

-21

2014

-jun-1

8

2014

-jul-

16

2014

-ago-1

3

2014

-sep

-10

2014

-oct

-08

2014

-nov-

05

2014

-dic

-03

2014

-dic

-31

Riga: Monthly calibrations 2001-2014

0

200

400

600

800

1000

1200

1400

1600

Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec