PS1 Seeing Observations:

A series of ongoing observations are being made explicitly to analyze the seeing conditions for the PS1 telescope. The first several nights of data were taken with the RoboDIMM located in the PS1 parking lot; on Dec. 9 it was moved to the PS1 entry balcony on level 1. Efforts were made to simultaneously observe both the same and different stars that the DNSM DIMM was observing. At times, we were also able to acquire simultaneous observation from the PS1 of the same star. Different symbols show which star was being observed. Different line colors separate the DIMM measurements. On February 5 the RoboDIMM was moved to the PS2 (formerly MAGNUM) dome.

PS1 RoboDIMM Measurements

PS1 Skin Seeing

Reber Circle Seeing

Faulkes Telescope North

PS1 'Fuzzball' Events

RoboDIMM Howto

CTIO DIMM documentation

PS1 RoboDIMM Measurements

December 3, 2009

RoboDIMM located on jibcrane pad in parking lot. RoboDIMM measurements worse than DNSM and GPC1 measurements for most of the night

December 7, 2009

RoboDIMM located on jibcrane pad in parking lot. RoboDIMM measurments were better than GPC1 measurements but worse than or equal to the DNSM measurements.

December 9, 2009

RoboDIMM located on L1 entry balcony. RoboDIMM measurements were worse than DNSM measurements but better than GPC1 measurements for most of the night. Winds from the N-NW.

December 11, 2009

RoboDIMM located on L1 entry balcony. Note that this night shows particularly good correspondence between the DNSM and RoboDIMMs seeing measurements. Both DIMMs show better seeing than the GPC1 measurements. Light tradewinds throughout the RoboDIMM measurement period

December 16, 2009

RoboDIMM located on L1 entry balcony. The wind during the RoboDIMM measurements was ~5 m/s out of the N-NW. During the second set of GPC1 images both the DNSMand RoboDIMM were running but not able to collect data.

December 24, 2009

RoboDIMM located on L1 entry balcony. The wind during the RoboDIMM measurements was light (<3 m/s) and varied in direction from W to NW. The target stars for both DIMMs were in the SE so data was collected pointing away from the wind.

December 28, 2009

RoboDIMM located on L1 entry balcony. RoboDIMM located on L1 entry balcony. The wind during the RoboDIMM measurements was light (<4 m/s) from the N-NW. The target stars for both DIMMs were in the SE so data was collected pointing away from the wind. GPC1 data is corrected for airmass.


From Dec. 28, 2009 onwards, DNSM data are corrected to zenith. Also, from this date, DNSM and RoboDIMM data are calibrated to PS1.

The format of the data files is:

time_UT time_msec name r0 zenith_distance fwhm_uncorr fwhm_corr fwhm2ps1
time_UT time_msec name_1 name_2 r0 fwhm fwhm2ps1

December 30, 2009

RoboDIMM located on L1 entry balcony. Wind from SW at < 2 m/s. Target stars in the SE.

December 31, 2009

RoboDIMM located on L1 entry balcony. Wind from W at < 3 m/s. Target stars in the SE.

January 3, 2010

DNSM not operating. RoboDIMM located on L1 entry balcony. Wind from W-SW at < 8 m/s. Target stars in the SE.

January 6, 2010

DNSM not operating. RoboDIMM located on L1 entry balcony. Wind from W at < 10 m/s. Target stars in the SE.

January 7, 2010

DNSM not operating. RoboDIMM located on L1 entry balcony. Wind from W at < 10 m/s. Target stars in the SE.

January 8, 2010

RoboDIMM located on L1 entry balcony. Wind from W at < 8 m/s. Target stars in the SE.

January 10, 2010

RoboDIMM located on L1 entry balcony. Wind from W at < 6 m/s. Target stars in the SE.

January 19, 2010

RoboDIMM located on L1 entry balcony. Wind from NNE at < 1 m/s. Target stars in E & SE

January 20, 2020

RoboDIMM located on L1 entry balcony. Wind from NW at < 1 m/s. Target stars in E & SE

January 21, 2010

RoboDIMM located on L1 entry balcony. Wind from N at ~ 5.5m/s. Target stars in E & SE. Last couple of minutes we lost the star due to it being eclipsed by PS1

January 22, 2010

RoboDIMM located on L1 entry balcony. No wind. Target stars in E & SE

January 23, 2010

RoboDIMM located on L1 entry balcony. Wind from S-SW at < 5 m/s. Additional tarp hung to block wind. Target stars in E & SE. Operations halted at 8 after the wind increased and moved to due W.

January 26, 2010

RoboDIMM located on L1 entry balcony. Wind from S at < 6 m/s, but dropped quickly to nearly nothing. By 6 it had moved to the SW at 7 m/s. Additional tarp hung to block wind. Target stars in E & SE.

January 27, 2010

RoboDIMM located on L1 entry balcony. Wind from S-SW at < 7-10 m/s. Additional tarp hung to block wind. Target stars in E & SE.

January 28, 2010

RoboDIMM located on L1 entry balcony. Wind from W-SW at < 7-8 m/s. Minimal wind buffeting of RoboDIMM. Target stars in E & SE. DNSM restarted at 5:45.

January 29, 2010

RoboDIMM located on L1 entry balcony. Lots of dew with some drops on L1 balcony. Observations ended early.

January 30, 2010

Poor weather; no RoboDIMM observations.

February 2, 2010

RoboDIMM located on L1 entry balcony. Wind from S. Weather/visibility deteriorated at 7:30. PS1 did not open.

February 3, 2010

Poor weather in early evening; no RoboDIMM observations.

February 4, 2010

RoboDIMM located on L1 entry balcony. 3-4 deg. C with 20-30+ mph winds from the NW. Problem with CCDOPS program from 7-9:30.


RoboDIMM moved to PS2 (formerly MAGNUM) dome. To cool dome to ambient the large blower is run for an hour with the slit open.

February 5, 2010

Wind from E at 6-9 m/s. Dome cooled to ambient by 6:30.

February 6, 2010

Wind from SW at 4-7 m/s; moved to W at 9. Ran blower test at 9:15.

February 9, 2010

A windy and cold but clear night.Winds were 9 to 12 m/s from the NE.

February 10, 2010

A warmer than usual night, steady 10 C with light winds (3m/s) from NE. Humidity 13%. A few thin cirrus clouds.

February 11, 2010

Another nice weather night. Winds 5 m/s from the NNE, 10% humidity and near 10 C throughout the observing period. Some thin cirrus around.

February 12, 2010

Clear, fairly warm (8-8.5 C), with wind from the N-NE at 3-6 m/s. Temperature stable in PS2 dome.

February 13, 2010

Clear, with light (3-5 m/s) NE winds. Colder tonight (5-6 C) and low (< 5%) RH.

February 16, 2010

A clear night with 4-6 m/s winds out of the NW. Locked onto Sirius at 0508 U.T. after recovering from a COM 1 port non-connection. Moved to Regulus at 0800.

February 17, 2010

A clear and cold night with northerly winds at 6 to 19 m/s sec. Starting tracking Sirius at 0515. It was a bit difficult keeping the star in the box, it wanted to drift to the lower left apparently due to wind turbulence. Moved to Regulus at 0810 but due to wind buffeting we could not acquire for more than a few minutes. Wind conditions get worse so we quit attempts at 0930.

February 18, 2010

No observations due to PS1 shutter issues.

February 19, 2010

Nice, clear evening with winds 2-6 m/s from the NW. Blower started at 4:40; shut off at 5:30. RoboDIMM started at 4:45, looking at Sirius. Moved to Regulus at 7:30.

February 20, 2010

Clear with no wind at 4:30. Around 5 the wind picked up to 5-6 m/s from the NE. 30 minutes later it dropped to 2-4 m/s from the N-NE where is stayed until 10. Blower started and PS2 dome opened at 2:45. Seeing extremely poor early in the night. RoboDIMM imaging Sirius which moved all over (and off) the display. The same star was spattered across the DNSM display. Moved to Regulus at 7:30. Seeing continued to be poor throughout the night.

February 23, 2010

A partly cloudy night (cirrus) with winds, 2-4 m/s second out of the NNW. No wind in the dome. Seeing Monitor for the observing period was 1.5 to .5 at end of shift. Locked onto Sirius at 0453. On to Regulus at 0745 to midnight.

February 24, 2010

A clear start to the night with just a few clouds low in the west and east. Winds 2 m/s, temp. is 8. The moon was close to RoboDIMM's first star, Sirius. Most of the time seeing was sub arc second despite increasing cloudiness. . Locked onto Sirius at 0500. To Regulus at 0720. Blocked by clouds for a short time.

February 25, 2010

A clear night with good seeing throughout Robo's tracking period. Winds were around 4 to 6 m/s from the wsw, turning nw. First acquisition on Sirius, drifted away. Restarted the system and locked on at 0515. On to Regulus at 0745.

February 26, 2010

Clear, warm (11 deg) evening with no wind at the beginning of the night. Later it picked up (<3 m/s) out of the NW. Pointed RoboDIMM at Sirius at 5 UT; moved to Regulus at 6:45; Spica at 9:30.

February 27, 2010

Clear, warm (11 deg), and no wind at 5 UT. Winds gradually picked up (<4 m/s) from the N-NW. Followed DNSM star list; closed down at 14 UT.

March 2, 2010

Shift started out with clear skies, winds from the NE at 2-3 m/s. Temp was 5 C with 3 % humidity. At midnight winds are around 6-8 m/s. DIMM seeing is not good.

Locked onto Sirius at 0455. On to Spica at 1005 On to Arcturus at 1130. The telescope did not recognize this star as pre-programmed # 147. Wind is affecting data collection now. At 1345 to Vega. The scope was having tracking problems on Arcturus which was near zenith. Retried Arcturus but still having near zenith tracking problems. A quiet night otherwise.

March 3, 2010

A start of shift, skies are clear, wind 4 to 6 m/s from the ENE. Humidity5%, temp 7. Winds decrease after midnight. Seeing 1" to 1.5" most of the night. Supposedly, seeing conditions (tonight and especially last night) were affected by a 100+ mph sub tropical jet in the free atmosphere.

Locked onto Sirius at 0654. On to Procyon at 1010. PS1 has been doing non stop engineering from start of shift to 1028. Onto Arcturus at 1110. Arcturus lost track at zenith (its keyhole) so onto Antares at 1355 to end of tracking period at 1533.

March 4, 2010

A variable weather night that started out with with mostly clear skies except for a little bit of cirrus. Wind were 3-6 m/s from the NE. The subtropical jet affects seeing again. Clouds thicken a lot around 0930. Humidity rises (70's), PS1 closes due to sudden drop of visibility. Clouds are nearby but lower in elevation. At 1050, a little fog blows through. Clouds, humidity and periods of fog end the night about an hour early as PS1 is closed again.

A problem at start up. The telescopes 18v power supply was dead. It was replaced with another power supply from Haleakala's Amateur Astronomer's (Rob) surplus parts department. Locked onto Sirius at 0515. Lost track once due to clouds. To Procyon at midnight. Clear skies again. To Arcturus at 1110. Due to tracking problems with Arcturus at zenith, moved to Antares at 1345. Humidity and clouds again. PS1 closes. RoboDIMM closes at 1420.

March 9, 2010

At start of shift gusty winds, up to 20 m/s from the NNE made donuts out of the RoboDIMM image. Due to the winds the Seeing Monitor is down. We will try though. Locked onto Sirius at 0510. Star reacquired numerous times as RoboDIMM can only hold star for 3 to 5 minutes. To Spica at 0800. The Seeing Monitor is still down for weather. Qtfocus indicates poor seeing. Moved to Arcturus at 0905. Too windy to acquire. Now just waiting for the winds to subside or call it a night. At midnight, no change in weather. Going home.

March 10, 2010

At the start of shift we have winds gusting into the 20's m/s, high humidity, and fog. Temp is 1.9 C. Rob added some lead weight to the Mead telescope to mitigate vibration in the case we do observe. It appears that some of RoboDIMM's bouncing is caused by the winds vibrating the dome floor. Uninterrupted high winds and fog continue through midnight. We're bailing. No data collected.

March 11, 2010

Poor weather; no RoboDIMM observations

March 12, 2010

Clear but windy (6-10 m/s with periods to ~15 m/s) night. DNSM was closed from 10-11 due to high winds. Little DNSM data throughout the night probably due to a combination of high wind and poor seeing. Followed DNSM star list Moved to Antares at 12:50 because Arcturus was directly overhead.

March 13, 2010

Clear night with moderate (<7 m/s) E-SE winds. Seeing reported by DNSM consistently mediocre. Followed DNSM star list from 5-13 UT.


"Skin seeing" is a degradation in seeing due to a temperature differential between ambient and the surface of a dome. White domes, while reflecting heat during the day, cool to below ambient during the night. This temperature differential creates turbulence above the dome slit. The potential for skin seeing at PS1 is being investigated with a temperature monitoring system and series of seeing measurements using two RoboDIMMs.

Skin Temperature Monitoring System

To investigate whether a skin seeing issue exists at PS1 a monitoring system has been installed to measure dome skin and adjacent ambient temperatures. The system consists of three Resistance Temperature Detectors (RTDs) and a RTD input module. The photo show the placement of the sensors. The first dangles in the end a piece of 3/4" PVC. The sensor is shaded but the bottom of the tube is cut out to allow air flow. The sensor is approximately 2' away from the dome. The second sensor is taped to the dome with aluminum tape. The third is held against the dome surface by a 1" x 1" x 1/2" piece of closed-cell foam epoxied to the surface. Sensors two and three have a coating of thermal conductive paste. The sensors are wired to an ADAM-6015 RTD input module through the unused vent 3. The input module is polled every minute by a script on Otis 7. The data is recorded in /export/home/otis/skintemp/data with file names yyyymmdd.skintemp.dat Sensor data collected on August 18, 2010 illustrate the fluctuation of skin temperatures above then below ambient as the day progresses. The tempering effect of foil covering is also shown in the plot.

Dual DIMM Measurements

The original RoboDIMM control computer was an older laptop running Windows 98. Despite the attached GPS receiver, system time often varied widely from the actual time due to the load imposed by the RoboDIMM program. To get measurements from two RoboDIMMs relatively close in time, two relatively modern laptops running Windows XP were built with all of the required software.

On May 28, 2010, two RoboDIMMs separated by a distance of about 10 feet collected simultaneous data on Polaris from the PS2 dome. A plot of this data indicates the two DIMMs are quite well correlated.

Dual RoboDIMMs

On August 25, 2010, two RoboDIMMs were set up south of PS1 and pointed at Polaris along the west side of the PS1 dome. The goal was to have one DIMM measure seeing through the boundary layer above the dome surface while the other DIMM measured along a path outside the boundary layer. To locate the DIMM measuring seeing near-dome the DIMM was moved east until the images were blocked by the dome then moved west until they were just clear. The other DIMM was placed 5' to the west.

During the time the data was collected the dome was basically stationary while the observer was taking pinhole data on the moon. Because the moon was rising the near-dome measurements were made off the surface of the rear shutter. Seeing was poor during while the data was collected. Also, because the RoboDIMMs were at ground level there was a significant ground seeing effect. The data seem to indicate that any degradation of seeing due to the skin seeing effect is insignificant compared with other factors.


A series of measurements were made to characterize the seeing in the vicinity of Reber Circle. One set of observations was undertaken to determine if there is significant variation in seeing across Reber Circle. A second set was made to compare seeing at the center of Reber Circle to that at the top of the Airglow building. For comparison RoboDIMM data was plotted against seeing data acquired by the Day-Night Seeing Monitor (DNSM). Plotted DNSM data has been corrected to zenith and converted from r0 to FWHM.

To measure the variation in seeing across Reber Circle a static RoboDIMM (Location A on map) was positioned just off the asphalt NE of Airglow. Simultaneous measurements were made with the static DIMM and a second RoboDIMM at a series of four locations approximately 8 feet apart in a N-NW line (Locations B-E on map).

September 2, 2010

Measurements were made at Locations A, B, and C. Winds all night were light, <3.5 m/s. Throughout the night both RoboDIMMs and the DNSM were looking at stars high in the sky. The DNSM was pointed at Altair all night. Both RoboDIMMs measured Antares during the data taking at Location B. Around 6:30 UT some high clouds moved in. I believe this accounts for the degradation in seeing and the lack of DNSM data points. Because the DIMMs, particularly the static one, were struggling to keep their stars on the focal plane both were moved to Altair which was further south.

September 7, 2010

Locations 2,3, and 4. During the RoboDIMM observations at locations A, C, D, and E the wind was out of the SE at 3-6 m/s as measured at the DNSM. After struggling for nearly an hour trying to get the RoboDIMMs to track Vega, a star nearly as close to zenith as Altair, the star the DNSM was tracking, the DIMMs were moved to Arcturus which was dropping to the west. The disparity in reported seeing values in the attached plot during this period appear to reflect the difference between the RoboDIMMs and the DNSM in the altitude of their respective target stars. Despite looking through quite a bit of atmosphere the RoboDIMMs seem to be fairly consistent with each other.

At 7:45 both RoboDIMMs were moved to Vega. The DNSM target star continued to be Altair. The lack of DNSM data points from 6:45-8:45 appear to be due to high clouds. The PS1 observer reported high cirrus during the evening. Even without a large number of DNSM data points the r0 values appear to coincide well with those reported by both RoboDIMMs. The two DIMMs were 32 ft apart during the final data set yet seem to agree fairly well. The offsets in values, particularly around 8:45, appear to be due to clock issues. Although the GPS time program is set to synch every minute, around this time a couple instances were noted where the clocks were nearly 2 minutes apart. For most of the time, however, the differences are less than 20 sec.

September 10, 2010

Data was collected from the RoboDIMMs separated by a distance of approximately 32' on Reber Circle (locations A and E). Winds were close to 6 m/s early in the evening but dropped to just over 2 m/s by the end. High clouds necessitated moving the DIMMs to a number of different targets as well as repeatedly restarting the program and re-acquiring stars. The last hour was spent tracking a star quite low in the sky

September 14, 2010

Data was collected again at locations A and E, approximately 32' apart on Reber Circle. The winds were ~10m/s out of the E-SE during the evening. The static DIMM (location A) was sheltered from the wind somewhat by the PS2 containers. This seems to account for the differences between the DIMMs between 6:30 and 7UT. At that time the DIMMs were moved from a Arcturus which was fairly low to Rasalhague which was closer to zenith. Both DIMMs were affected by the wind but the one nearer the center of the Circle (location E) was buffeted more. Despite this, it appears that the measurements coincide well and it seems that we can conclude that seeing is relatively consistent across Reber Circle.

September 23, 2010

Airglow roof and Reber Circle center

Plot only has short time segment where "fuzzball" event didn't occur

October 14, 2010

Data was collected from the roof of the Airglow building (location F) and the approximate center of Reber Circle (location G). In the plot the DNSM data is corrected to zenith and converted to FWHM, the RoboDIMM data is converted to FWHM, and the PS1 FWHM is corrected for airmass. Throughout the night the winds were < 5 m/s out of the NE. The large gap in the DNSM data is due to high clouds obscuring Altair. At 8 UT the DNSM moved to Diphda and resumed data collection. The high clouds moved across the RoboDIMM's path to Vega a few times requiring that the DIMMs be restarted. During the time period that RoboDIMM data was collected PS1 was pointed to the SE.


On November 14, 2010, a RoboDIMM was set up at Faulkes Telescope North (FTN) to correlate seeing data with that collected by the Day-Night Seeing Monitor (DNSM).

November 14, 2010

RoboDIMM located at pivot point of FTN enclosure. Closed due to humidity.

November 23, 2010

RoboDIMM located at pivot point of FTN enclosure. Closed due to high winds, clouds.

November 27, 2010

RoboDIMM located on floor of FTN enclosure. Closed due to high winds, clouds.

December 5, 2010

RoboDIMM located on floor of FTN enclosure. Seeing too poor for DNSM to acquire data.

December 7, 2010

RoboDIMM located on floor of FTN enclosure. Variations in seeing throughout the night correlated well.


Sudden increases in the FWHM of PS1 images, or 'Fuzzball' events, are being investigated. Attempts are being made to correlate these events with environmental conditions to determine if a link exists.




  • Adjust tarps for anticipated wind (there's an extra in the wood chest)
  • Remove RoboDIMM cover *CAREFULLY* so as not to dislodge finder (remember carabiner at bottom)
  • Check to see that the leveling bubble is still mostly within the circle (close is okay)
  • Retrieve telescope cables, camera controller, power strip, keypar, and stool from L1


  • Release alt brake, flip telescope, attach keypad, RS232 connector, and 18vDC pwr (tuck all cables under tripod)


  • Set controller on stool, attach camera cable *BEFORE powering camera controller*


  • Set on camera controller, attach white cable from controller to LPT1, and serial from telescope (tape these together)
  • Plug GPS in USB port; set receiver in the open
  • Turn on power strip (then switch telescope ON and plug in controller)
  • Use down arrow and hit return to select "Windows" at GRUB (Suse) prompt
  • Press cancel at login screen
  • Double click NMEATime icon
    • Click the GPS Signal Quality tab (bars must be blue; GPS Locked LOCKED)
    • Click Update PC Clock Now
    • Close NMEATime window

Telescope Alignment

  • Get a star in the finder (there are 4 speed buttons on the keypad)
  • With the keypad:
    • Use down arrow to select "Align" press Enter
    • ALTAZ press Enter
    • 1 Star press Enter
    • Level base (should be) press ENTER
    • press ENTER, then scroll (with arrows) to align star (probably Aldebaran)
    • In the flip mirror viewport the star should be aligned so the lower image is centered on the crosshairs of reticulated eypiece
    • Turn the flip mirror knob

Running SBIG software (CCDOPS)

  • Double click the CCDOPS icon
  • There will be a message "Trying to Establish Link with Camera"
  • If a link cannot be established, unplug controller, gently reseat all camera connections, power up and try again.
  • Click on the Focus button; a dialog window ill come up with and exposure time of 0.25 sec. This is appropriate for the bright stars we're currnetly using; click okay
  • If the force is with you you'll get a double image centered in the field
  • If not, use the keypad on the lowest speed (bottom button) to adjust their position. (It does not have to be perfect.)
  • No image? Check the flip mirror.
  • Close CCDOPS (if there are messages about cooling ignore them)

Running RoboDIMM

  • Click the Newtest_robodimm2.exe icon (looks like a dome)
  • Check Auto Measure Enable checkbox
  • Click Expose CCD button
  • If there is a message about a telescope error, reseat the telescope connection to the laptop (the serial port). You may also have to cycle power to the telescope (and repeat the alignment).
  • Confirming Operation:
    • If there are two spots inside a flashing white box the odds are RoboDIMM is happy
    • In the lower left of the window there will be output:
      • Last Seeing Data should have a time that's fairly recent
      • dx ~= 90 and dy ~= 0 (These values refer to the orientation of the prism and should not change unless the cover is dislodged. If it is, adjst dy by loosening the thumbscrews and rotating the cover.)
      • Last Strehl Ratio for Left/Right? ~= 0.5
    • Leave the platform gently


  • Record wind speed, direction, and ops start time
  • Check for a flashing box/star images ~10 min intervals

End of Night

  • Click Quit button
  • Turn telescope OFF and unplug controller
  • Telescope: unplug keypad, RS232, and 18 vDC
    • Release alt brake, tip so camera is on top, wrap cord carefully
    • Bag carefully; lightly cinch straps opposite finder; clip carabiner
  • Laptop: pull serial, unplug LPT1; check the time to make sure it's not way off
  • Controller: unscrew camera (black cable)
  • Data is in C:\dimmdata2002\D1test_seeing_lab.log
    • Open with double click, hightlight and copy the night's data; paste into a notepad file (icon in lower bar)
    • Save as MM-DD.txt (maybe leave off 0's because win98 is a bit dim) on F: (usb memory stick)

Last modified 10 years ago Last modified on Jan 19, 2011, 11:16:08 AM

Attachments (138)