Optimized flight planning If for SOFIA

J. M. M. Horn, and E. E. Becklin

University of California, Los Angeles,� Division of Astronomy and Astrophysics�

405 Hilgard Avenue, Los Angeles, CA 90095-1562�

Optimized flight planning for SOFIA

J. M. M. Horn and E. E. Becklin

University of California, Los Angeles, Division of Astronomy and Astrophysics 405 Hilgard Avenue, Los Angeles, CA 90095-1562

ABSTRACT

A study of the average upper atmospheric conditions has been carried out in order to optimize the scientific return from SOFIA.. By examination of atmospheric data from satellite missions, we found that at typical SOFIA flight altitudes (between 37,000 and 45,000 ft), it can be an advantage to fly north, as the water vapor overburden (WVO) and the frequency of cloud occurrence is less than if the flights where centered above Moffett Field, CA, which will be the base for SOFIA. It has also been shown that for certain science projects, the amount of time on target can be considerably extended. An increase of up to 50 % of target time compared to typical KAO flights of 6.5 to 8 hours, is possible without exceeding the often required very low line-of-sight (LOS) WVO of;:::; lOJLm. This converts into a significant increase in observing efficiency for SOFIA. Because SOFIA is flown by United Airlines, only military territories are potentially restricted areas. So flights out of Moffett Field, CA, going into Canada have been studied as well as up to week-long flight series in and out of Anchorage, AL, flying over Canada and Alaska, where the air is very dry at SOFIA altitudes in the winter months. Those flights have shown that the LOS WVO is almost always below lOJLm. This is true even for flights that start at 37,000 ft and thus produce extremely long flights up to 12 hours. Another impo.rtant result of this study is that it shows the LOS WVO for a fixed set of galactic plane objects is about the same for north shifted SOFIA flights, elevation range 20-60°, as for the KAO, elevation range 35-75°.

Keywords: SOFIA, flight planning, atmospheric transmission, airborne astronomy, water vapor overburden

1. INTRODUCTION

Between 30/lm and 300 JLm, the water absorption in the atmosphere is so strong, it essentially prohibits any obser­vations from ground. Between 4 and 600 JLm there are only relatively few windows which also depend on water vapor content. Flying at altitudes of typically 41,000 ft, the Stratospheric Observatory for Infrared Astronomy, SOFIA, flying on a Boing 747SP, will be operating above 99 % of the atmosphere's water content, leaving between 5 and 10 /lm water vapor column above it. The mid infrared region from about 30 to 300 microns therefore becomes about 80 % transparent, with some water absorption lines and almost spaceborne conditions between them. l

SOFIA has an elevation range of 20° to 60°, and a cross-elevation range of ±3°. An astronomical object that shall be observed therefore imposes constraints on the flight direction as the telescope can virtually not be moved in azimuth. So observation planning on SOFIA inheritely means flightplanning too. As long as one object is observed, the flight leg is predetermined and fixed. As soon as the object is out of SOFIA's elevation range, another object has to be observed, which usually means that the flight direction has to be changed too. So a flightplan looks pretty much like a zig-zag, and furthermore, the aircraft typically has to land at the same location where it departed.

Rather than flying only at a single altitude, e. g. 41,000 ft, SOFIA will often start at lower altitudes and climb to higher altitudes as more and more fuel is burned away. Some flight profiles with various total flight times calculated by United Airlines2 are shown in Fig. 1 and demonstrate how such flight profiles could look like. Those flight profiles are specifically calculated for the SOFIA aircraft, a Boing 747SP, at the expected weight and drag. They depend on the zero fuel weight (ZFW), which was assumed to be 370,000 lbs for the calculations. The latest number for the ZFW as of January 2000 is 374,000 Ibs.3 Qualitatively, the flight profiles for KAO flights looked similar. Parameters like aircraft weight or Mach-Number were different on the KAO, and especially the maximum flight time was shorter. However, step profiles between 37,000 ft and 45,000 ft as e. g. in flight profile #1 of Fig. 1 were very common on the KAO. Very often, observations started at 37,000 ft and 39,000 ft. 4

Further author information: (Send correspondence to J. M. M. Horn) J. M. M. Horn: E-mail: horn@astro.ucla.edu E. E. Becklin: E-mail: becklin@astro.ucla.edu

To appear in Proc. SPIE 4014, 2000 1

SOFIA (747SP) Flight Profiles Initial Time on Total Gros, Weight Altitude Flight Time

659.600 lb' 11.0h 12.2 b 1. ~

FUIO

2. ~ 11.6h 12.6h

FL370

3. / tUh \ 11.5 h 12.4h

FUIO fL430 593.800 Ib, 8.5 b 9.5 b

4. ~2.5h 2.5h 3.5h FUIO

5. / FL390·

2.6h 6.5h \ 9.t b IO.Ob

FlA30 FL45O

6. ~ 8.8 h 1O.0h

5.1h 2.5h 1.2h FL390

7. / 9.111 9.1 h 9.9 h

FlArO

8. f 6.5h \ 534.600 Ib, 6.5 h 7.3 h

9.

FlArO FU30

e-:::\ 6.0h 6.9 b

Figure 1. various flight profiles possible for SOFIA. The calculations hav'e been performed by U.4.L, except profiles 1, 4 and 9, which have been produced by the author using U.4.L tables. Indicated are the flight levels in hundreds of feet (FL41O: Flight l~ve141000 f~t), and the corresponding cruise times on each of these levels in hours. Additionally the initial gross weight of the aircraft is given, indicating the increased fuel load necessary for longer flights.

Flights with the KAO had a typical duration of about 7-8 hours, and were centered slightly north-east of Moffett Field, CA, the KAO base and also the future SOFIA base, see Fig. 2. In this study, the extension of the overall flight time to about 9-10 hours in order to increase the science return of the observatory has been investigated. From the flight profiles discussed in Fig. 1, one can see that a flight time extension, makes it necessary to load more fuel and therefore to start flying (and observing) at lower altitudes, e. g. sometimes at 37,000 ft. Often it seems attractive to observe from as high an altitude as possible for several reasons. First the zenith water vapor overburden (WVO) is lower. The amount of water is crucial e. g. for spectroscopic line observations in the vicinity of atmospheric water absorption lines. Second there are high-altitude cirrus clouds which affect science especially in the near and mid infrared region as the typical particle size of the ice crystals is about 10 J.Lm. On the other hand, as already mentioned, it is not the zenith WVO alone that counts. Also the object's elevation or, both parameters combined to yield a line-of-sight (LOS) WVO has to be considered. Shifting many flightplans to northern latitudes, where the hygropause is expected to be lower than towards the equator,6 we demonstrate that the average LOS-WVO is reasonably low, so that science projects should not obviously be affected.

One of the observatory's contractial requirements is to achieve 960 successful flight hours (SFH) per year. SFHs are defined as observing time during which the zenith WVO is less than 10 J.Lm and the observatory is fully functional.

The zenith WVO limit does not take into account that many science projects can be very efficient even if that overburden limit is exceeded.7 We show in this study, that actually the LOS-WVO is much more useful, as it takes flightplanning into account. That is the object's elevation has to be considered as well as the amount of water in the atmosphere at the location of the flightleg. With an average elevation of 400 for SOFIA, the SFH-requirement of lOJ.Lm zenith WVO translates to 15.6J.Lm LOS-WVO. By looking at many flights, so that statistical conclusions on the flight legs can be drawn, one can say that on average any observation time where the LOS-WVO is above the 15.6 J.Lm adds to 960 SFH a year. Possible telescope or system downtime also adds to the 960 hours.

The amount of water vapor above certain SOFIA flight levels was available on a global distribution from the

To appear in Proc. SPIE 4014,2000 2

48 KAO Research Flights. 1993 - 1995

Prlnt

A. Mever. Nov. 1995

Figure 2. The distribution of 48 KAO research flights out of Moffett Field, CA, in the years 1993-1995. Each turning point in a flightplan marks the end or start of an observing leg. Besides the borders of US states, restricted areas are shawl} too, as e. g. the boxes on the Pacific West of southern California5

Microwave Limb Sounder (MLS) on the Upper Atmospheric Research Satellite (UARS). On the same satellite, the Cryogenic Limb Array Etalon Spectrometer, CLAES, simultaneously measured the frequency of cloud occurence.8

Therefore, an analysis of the clouds above SOFIA as well as the amount of precipitable water was possible. In a high-altitude site survey, Haas & Pfister (1998)9 used MLS-data from a previous MLS-retrieval to get information on which air-base provides the best atmospheric conditions at typical SOFIA flight levels. Flight time extension or north-shifted flight plans were not considered.

By making use of the lower telescope elevation range of SOFIA (200 -60°) compared to the KAO (350 -75°), SOFIA is able to observe an object lying in the southern sky, e. g. Sgr A·, from farther north than the KAO. By monitoring the MLS water vapor data, we can show that the science is not compromised in that particular example. The same results have been found with other KAO flights, observing a variety of objects on flights out of Moffett Field. This case as well as the study of a flight series out of Anchorage, AL, with an expected very low WVO, will be discussed briefly in Sect. 3.2 and Sect. 3.3.

We also present a statistical study of 16 flights out of Moffett Field that have been created, using objects which from previous observations with the KAO and some new objects from ISO observations. Those flights have been grouped in 4 subsequent days, each 3 months apart, to cover the different seasons. By systematically shifting those flights to northern latitudes, we show that on average, the SFH water vapor limit is exceeded for only 3 % of the observing time.

The ultimate goal, which goes beyond this study, is to get the satellite data as early as possible after they have been taken in order to generate water vapor forecasts. If one is able to generate reliable forecasts, the flight planning could be substantially improved. As a precondition, the day-to-day variations need to show a history, which requires good data and a good model. For this particular purpose, the spatial sampling of MLS on UARS with 1300 points per day, is too low. The day-to-day variations therefore don't allow for a reliable forecast at most they indicate how isolated regions of large amounts of water vapor move from day to day. Furthermore, MLS is no longer regularly operating, so a constant stream of data is no longer available. However, there is a follow-up mission, the MLS on the Earth Observing Satellite (EOS), which is similar to the MLS on UARS, but has better spatial sampling with 3600

To appear in Proc. SPIE 4014, 2000 3

vertical profiles each day, and the vertical resolution is better too with about 1.0 km compared to 3 km with MLS on UARS. The launch date of MLS on EOS is late 2002. With this instrument, the approach to getting forecasts is very promising. The data might be available two days in advance. Together with other upper atmospheric weather data like wind directions and velocities or clould occurences, water vapor forcasts might in fact be useful for SOFIA operations.

In Sect. 2 we briefly describe the model we use to get the water vapor overburden out of the satellite data. Also a validation of the model used is presented. Thereafter, in Sect.3, the north-shifted SOFIA flight plans are shown, as well as the flight series out of Alaska, and the comparison of the Sgr A· observation from a KAO flight and a north-shifted SOFIA flight. In the conclusions we finally review the results and discuss the impact of the modified flightplanning on the SOFIA operations.

2. THE DATA

The MLS on UARS measures the relative humidity for four different pressure altitudes hhPa, ranging from 6.2 km to 13.8 km. From these humidities, then a water vapor mixing ratio q in [parts per million volume - ppmv] is retrieved. The spatial sampling is about 1300 data points per UT-day and the latitude coverage switches about every month from 80S-30N to 30S-80N. 10

Former satellite and radio-sonde missions have accurately determined the vertical profile of the water vapor mixing ratio. It increases exponentially until a certain height hi, the hygropause. For higher altitudes h, it stays constant at about 4 ppmv9 •11 :

-h for h < hi,q(h) = ~O' exp ~ (1) { for h ~ hi

20

E 18

~

I:i 16 12 ::> -e 6

14

8­;: 12

u 10

~ -5'c

8

~

] 6

-If 4

'" 0 U 2

0 0 20

MLS--<lerived Zenith Warer Vapor Overburden [I"m!

Figure 3. For three different flight series, the zenith water vapor overburden calculated from MLS satellite data are compared with the zenith water vapor overburden (zWVO) derived from spectroscopic line measurements with the cryogenic grating spectrometer (cgs) on the KAO. The symbols indicate different KAO flight-series. Scattering and error bars are discussed in the text.

To appear in Proc. SPIE 4014, 2000 4

with

h' = -ho ·In 4/qo,� (2)

and ho =1.3 km, the scale height of the water vapor mixing ratio as has been derived from the SAGE n· experiment, and qo the mixing ratio at the ground.

While orbiting the earth, MLS measured q(h) for a certain geographical location. Therefore we can integrate the profile to get the amount of precipitable water vapor overburden in microns for a particular flight level hFL:

00

z(h) = :H2~ . Jp(h) q(h) dh,� (3) dry aIr

h FL

with p(h) the air density for the upper atmosphere, MH20 is the molecular weight of water, and Mdry air the molecular weight of dry air.

For the purposes of this study, the model has been validated by checking it with independently measured zenith WVOs from astronomical observations. Those measurements have been performed on the KAO with the cryogenic grating spectrometer (cgs). 13

Fig. 2, which shows the correlation between the cgs-derived and the MLS-derived zenith WVOs, has been generated from 11 cgs flights on the KAO between January 1992 and May 1993. On some legs of those flights, at various altitudes, the water vapor overburden has been analyzed from spectroscopic 0 I data @ 63 j.£m.4.14 The position of the cgs measurement on the flight was noted on the MLS map for that day and the LOS-WVO was compared.

Some uncertainty" though still persists, as the MLS data are taken with a mix of day and night measurements and reflect an entire UT-day, whereas the flights are usually only scheduled for night-time, so e. g. the first half UT-day. So in a worst case, the first flight legs in the early UT-day are compared with MLS data taken at the end of the same UT-day. ,

Summarized, there are six contributors to the error bars and scattering of the points in that plot:

•� The scale height of the water vapor mixing ratio has been assumed to be constant; it will though rather vary with latitude and longitude

•� The vertical profile of the water vapor mixing ratio has been assumed to be constant at 4 ppmv above the hygropause, which is strictly not correct

•� The uncertainty of the time when the MLS measurement took place

•� The errors in the MLS measurements

•� The errors in the eGS measurements

•� The uncertainties in the location of a cgs-measurement

The first four error sources lead to the scattering of the points. The last two have been considered by the error bars. The errors in the eGS measurement have been estimated to ±1 t-tm zenith water vapor by.14 Often, the longitude and latitude where the measurement took place could not be determined very accurately. Sometimes only the leg at which the cgs-measurement took place was known, so that it was necessary to average the MLS-WVO value along the leg. For that reason, sometimes large error bars occur for the MLS-WYO. We feel that the correlation presented in Fig. 2 demonstrates clear the above method and the reliability of the error estimates.

·Stratospheric Aerosol and Gas Experiment III2

To appear in Pmc. SPIE 4014,2000 5

9' 3 RM TUE.,21 DEC·. 1999 fLTSTRT.)RN.RSC SCALE = l:Z .OZE'07

Figure 4. All the flightplans from the statistical investigation of north-shifted flights. Also shown are the restricted areas over the US, which still have to be avoided.

3. SOFIA FLIGHTPLANS

.3.1. Flight Series out of Moffett Field

For statistical purposes, four flight series out of Moffett Field, CA, have been studied, each series taking place at a different time of the year. The 16 flights have been pushed as far north as possible without exceeding the ma.ximum flight time, which is at the moment about 12 hours, see Fig. I. In particular, the dates have been chosen corresponding to the MLS data available, but attention has been taken to try to get 4 subsequent days for each flight series, as this will probably be the usual way of SOFIA operations. Only flightprofiles from Fig. 1 have been chosen, with a varying total flight time from 7 to 11 hours, or observing times of about 6 to 10 hours, respectively. Their track over continental North America is shown in Fig.4. The average flight time in this sample of 16 flights is 9:10 h and the average observing time is 8:00 h. Although not strictly necessary the flights are usually only during night time. Therefore as seen in Fig.5 only 2 flights have one leg partially after sunrise. Especially at longer wavelengths, e. g. above 50 microns, day-time observations are possible due to the lately improved performance of gyro-tracking systems.

By making the best use of the available flight profiles, the lowest reasonable altitude, FL370, had only been used for four legs of the entire 16 flights. The risk of having higher WVO due to the necessity of flying at those lower altitudes is therefore strongly reduced. Nevertheless, the line-of-sight WVO is still within the range that can also be found in the somewhat higher flight level on 39,000 ft, and even on 41,000 ft. This can nicely be seen in the histogram plots of Fig. 6. From that figure, one can also read off that the observing time above the successful flight hour (SFH) limit of 15.6 J1. LOS-WVO, which is the equivalent of the 10 J1.m zenith WVO for an average elevation of 40°, is 4 hours. Those 4 hours make up 3 % of the total observing time which is 128 hours.

With only 3 % of the total time exceeding the water vapor limit, we would therefore need to have 990 observing hours. Any other failures during the flight, prohibiting observations, add to this time. The maximum potential flight time per year is approx 1400 hours, given by the fact that the observatory will operate 44 weeks per year with four 8-hour flights per week.

By looking at a sample of MLS-maps (Oct., 3/4 1992) from this statistical study shown at the end of this paper, one can also track the day-to-day variations, see Figs. 9. The MLS-maps are though not ideal for such purposes. First, the spatial sampling is too low. An MLS data point reflects a vertical water vapor profile for a certain geographical location, or latitude/longitude parameter pair. There are only about 1300 of such vertical profiles every

To appear in Proc. SPIE 4014, 2000 6

E 182: c� ---- Jan. 01. 1992

__________________ ---. Jan. 02. 1992 _] 16

.-.-. Jan. 03. 1992 of 14� ~ -.-- Jan. 05. 1992

~ 12� ~ 8. -> 10

&>:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00

UT

C. Sununer

- Jul. 20.1992 _____� --_. Jul. 21. 1992

._.-. Jul. 22. 1992 -.-- Jut. 23. 1992

'§!l!

u ~

; 8

.. 6 iii J. 4 9 ~ 2 :J

&>:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 UT

B. Spring ~Or-~..,.~_.~_____.~~.,_-r-...,.___~,__~_.~___,

~ 18 c ---- Mar. 19.1992 ] 16 ---- Mar. 20.1992

------------------ .-.-. Mar. 21, 1992 u

of 14 ~ _.-- Mar. 22. 1992 .§ CIl~ 12 ~ ~

8. -> 10 u

~ 8

~ 6 en

~ 4 I

:J ~ 2� l

&>:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00

UT

D. Fall 20 r-.,.-.,----.~-.~---rr=========o'i,...___,

---- Oct. I. 1992 ]; \8 i! ---- Oct. 2. 1992

i ~ Oct. 3.1992c: i 1

] 16 ------t~-,-- Oct. 4. 1992 I'-----'----'.u

1) 14 U ! I .1: :- g_.~ -'~f1 c

<2 12 i l_.�

~

I8. r--I-> 10 t t..... "I

u :: !l ; 8� I --I ~: i!~ I I

L_~-,~, ! ~ !:.~L..r l_,,;.'.:,_~ 6

iii ~ 4

I

::! 2 :J

&>:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 UT

Figure 5. The Line-Of-Sight WVO for 16 flights over one year out of Moffett Field, CA, the future base of SOFIA. Each plateau in the curves shows one flight leg with the corresponding WYO. Plotting this vs. UT allows the evaluation of the length of each leg and the total observation time. The initial climb and descent are not plotted. The dashed horizontal line marks the successful flight hour limit of 15.6 J.Lm for the average elevation of 4C?

day. Also, those profiles are only probed at 4 altitudes, due to the limited vertical resolution of the instrument. And furthermore, only 2 of those points on the vertical profile have been used here, due to wet and dry biases of the other ones (see also Sect. 2). But still, given those limitations, one can actually track regions of high zenith WVO over the continental US. The westward movement of the region with a large amount of water vapor over Canada in Figs. 9 at 37,000 ft (FL370) is to be expected, as it goes in hand with the jet stream on the northern hemisphere at those latitudes. 15

Also care has been taken in order to cover a reasonable range of the sky from Moffett Field. Nevertheless, SOFIA will also operate regularly from New Zealand for a couple of months, so that observations of southern objects is no issue.

To appear in Proc. SPIE 4014, 2000 7

A. FL370 B.FL390

<LOS-WVO> '" 9.9 11m (<zenith-WVO> '" 6.4 11m) Total # of Legs: 34

2 4 6 8 10 12 8 10 12 14 16 18 20

C. FLAIO D. FLA30

2 4 6 8 10 12 14 16 18 20 2 4 6 8 10 12 14 16 18 20

Line-Or-Sight WYO [Ilm] Line-Or-Sight WYO [j.lm]

Figure 6. Histogram plots for all legs of the 16 flights out of Moffett Field, showing the number of occurences, or legs, of certain amounts of LOS WVO for the various flight levels. The histogram bin-width is 1 J.Lm.

3.2. Flight Series out of Alaska

As SOFIA will be flown by United Airlines (VAL), virtually all VAL destinations are accessible for SOFIA. So in one scenario, we created a flight series from Moffett Field, CA, (MoFi in the table below) to Anchorage (Anch), AL, and back, with a couple of flights out of Anchorage. Some statistics of that flight series is shown in the table below. And as expected, the WVO is very low and within the SFH limit. .

Flight- and performance data of the Alaska flight series Altitude Time on Altitude (L05-WYO) [J.Lm] Flight #1: 01/01/92, MoFi ~ Anch FL370 N. A. N. A. FL390 2:30 h 8.8 FL410 6:30 h 7.1 Flight #2: 01/02/92, Anch ~ Anch FL370 N.A. N.A. FL390 2:15 h 6.9 FL410 6:30 h 7.8 Flight #3: 01/03/92, Anch ~ Anch FL370 1:20 h 14.3 FL390 2:40 h 8.7 FL410 6:30 h 9.3 Flight #4: 01/05/92, Anch ~ Anch FL390 1:42 h 7.8 FL410 6:30h 8.7

To appear in Proc. SPIE 4014, 2000

Time on Altitude (L05-WYO) [J.Lm] Flight #5: 01/06/92, Anch ~ Anch 2:30h 9.7 2:30h 9.0 6:30h 7.8 Flight #6: 01/01/92, Anch ~ Anch 0:50 h Dead Leg 2:30h 9.6 6:30h 10.0 Flight #1: 01/08/92, Anch ~ MoFi N.A. N.A. 2:00 h 8.3 6:30 h 11.0

8

• • • • • • • • •

• •

l8 ~ln;al K.AO 11lgN .J.:I1c: JW'It 17'·. 1m •

LS WVO \'M~ hom J-d.y ....C'l""...lJ"t: July 2(}'22. 1992

.. SIfA Srr A

~ .. • • • • Srr A • -:.. 12 Mill r.7K J A ~ 12 Mill G18.l Srr A Solum v; ;r.• Srr Srr A•• • •

w.'1 . ~ Srr A8 • • • • $ 8 • • • •Srr A~ ~ WI'CED:!ll'.:.:; , :.:; NCCl9JO CEDlO' •I •I NGC 13)'

NCo{' '<9JO SaNrn HDlfll7% HO 161796

Flil"" Leg"

Figure 7. The comparison of the LOS- WVO of a Sgr A· flightplan on the KAO at latitudes of about 20N and a north-shifted SOFIA Sgr A· Bightplan at latitudes of about 40N. The LOS- WVO values have been calculated by overlaying these flightplans onto a zenith water vapor MLS-map, and taking the airmass into account. As no MLS-data were available for the original flight date, June 17 1992, a three day average of the most recent available data before the Bight, May 28-30 1992 (left), and the most recent data after the Bight, July 20-22 1992 (right) , have been taken.

3.3. Comparison of KAO and SOFIA flights

One interesting point which has to be considered when shifting a flightplan into northern regions is the observation of a galactic plane object. The trade off one faces is to fly with SOFIA out of Moffett Field~nto southern latitudes and observe the object at higher elevation, or fly further north and observe the same object at lower average elevations. The question is for which case is the actual LOS-WVO higher? There was such a flight on the KAO, where SgrA· has been observed for about 3:20 hours, flying south out of Moffett Field to latitudes of about 20N (approximately Hawaii), so that the object is at elevations between 35° and 40°. The flight profiles were available from KAO housekeeping data, which are filed in a cgs archive at NASA Ames Research Center, Moffett Field, CA.4,14 The flight plans are a direct product from the KAO flightplanning tool knav. On the north-shifted flight, the object has been observed at elevation angles between 20° and 25° from latitudes of about 40N, for a total of 4:00 hours.

As one can see from those two figures is that the WVO varies by about 20 J.Lm for the two different periods with respect to one flightplan, and that once the southern KAO flightplan would be more favorable and once the northern SOFIA flightplan. Note however that due to the longer flight time with the north-shifted flight additional objects from a list of popular KAO objects by Meyer5 have been scheduled, for which therefore no WVO data in the KAO flight exist. One can conclude from this that the modified ftightplan is still a gain, not only because additional objects can be observed at a reasonable LOS-WVO below the SFH limit of LOS-WVO =15.6 J.Lm. Also the KAO flight, if scheduled in the July 20th to 22nd period would entirely fail the SFH limit requirement, see Fig. 7.

Two other examples of the LOS-WVO comparison between north-shifted and orignal, south-bound KAO flights are shown in Fig. 8. These comparisons demonstrate that we get longer flights with SOFIA than with the KAO without compromising the LOS-WVO and thus science.

4. CONCLUSION

Increasing the flight time and therefore having more observation time on lower altitudes, like e. g. 37,000 ft and 39,000 ft, only slightly affects the water vapor overburden if they are shifted north over Canada, or if flying out of Anchorage, AL. For the north-shifted flight series out of Moffett Field presented in Sect. 3.1, on average only 3 % of the observing time exceeded the maximum water vapor overburden limit. The flightplanning method presented here therefore adds only little extra flight time in order to achieve the required number of successful flight hours of 960.

The study of 16 flights out of Moffett Field, resulted in an average observing time per flight of 8:00 h, or 9: 10 h total flight time. Most observations are taken at a flight level of 41,000 ft. Strictly going to 43,000 ft as soon as

To appear in Proc. SPIE 4014, 2000 9

32.0 r----,---,r--r-----,--r-----,-...,----,-...,----,--,---,

Orisia:&J KAO nighl doJ,le: J:an. 10'" 1992 28 28.0 _ KAO nish l I

I • SOAA nish' I ~ 24 24.0

i : 20 ~200 o >

b~ .. M42> 16 :t 16.0

M42~ t i •

M42of. 12 ~ 12.0 NGC2023IC 10 IC 10v; IC446 NGC 2146 NGC 2146 NGC 2071 HH7-11 NGC 4214, ::J • • • W3 • HH7-11. NGC 4449_ MR:~~K848 ~

~ 8 8.0 ~

a••

• •• •• - : . : . .. .::J

4.0

Aish' Lea'

Figure 8. The L05-WVO comparison between two KAO Bights and their north-shifted versions. KAO flight data were January, ~h and l(fh 1992, where also the MLS-maps were available.

enough fuel is burned out and the aircraft is light enough can bring quite some improvement. This is particularly true, if observations strongly depend on low water vapor overburden, such as emission-line spectroscopy in close vicinity to atmospheric water absorption lines.7

Finally we can say that the ultimate goal of this study is to get the Water Vapor and further informations on paramete;-s ~uch as upper atmospheric winds, and clouds, early enough to predict forecasts. DeJ)ending on the quality of the data available, one-day- or two-day-in-advance WVO-maps must be on hand. The next generation Microwave Limb Sounder on the Earth Observing Satellite (EOS) can probably provide this. 16 This instrument will be launched in 2002 and has a much better vertical resolution of 1 km from 6 km altitude up to 60 km. The sampling is about 3 times higher than with the current MLS-data, about 3600 data points. The hygropause can then be exactly determined, resulting in a much better model then presented here. Nevertheless, the existing MLS (UARS) data still provide a precious data base in addition to the forecast data that could be built into the observation planning tools of the SOFIA Data Cycle System. This is at least true for the first couple of years of operation, until a similar data base has accumulated with the MLS on EOS. Weekly or monthly averages can be generated for the same periods as an observation is planned to look where it is actually most advantageous to fly to. Similarly, with an older retrieval of MLS data, seasonal averages over the continental US have been generated for the years 1992-1994.9

ACKNOWLEDGMENTS

We like to thank W. G. Read from NASA-JPL for providing the MLS-data and for many helpful discussions, M. Haas from NASA-ARC for providing the flight- and water vapor data from the cryogenic grating scpectrometer (cgs), J. Mergenthaler from Lockhead for providing the CLAES-data, Brian Chapman from UAL for providing the calculations on flight profiles and Allan Meyer for providing many KAO flight data.

REFERENCES 1. E. F. Erickson, "SOFIA: The next generation airborne observatory," Space Sci. Reviews 74, pp. 91-100, 1995. 2. B. Chapman, United Airlines, SFO, priv. comm., 1998. 3. J. McCoury , USRA, priv. comm., 2000. 4. M. R. Haas , NASA ARC, priv. comm., 1998. 5. A. Meyer, NASA ARC, priv. comm., 1998. 6. L. Pfister, NASA ARC, priv. comm., 1999.

To appear in Proc. SPIE 4014, 2000 10

7.� E. F. Erickson, "Effects of Telluric Water Vapor on Airborne IR Observations," P P 110, pp. 1098-1105, 1998.

8.� J. 1. Mergenthaler, A. E. Roche, J. B. Kumer, and G. A. Ely, "Cryogenic Limb Array Etalon Spectrometer observations of tropical cirrus," ~ h s. Res 4, p.p 84-2219.1,199 .

9.� M. R. Haas and 1. Pfister, "A High-Altitude Survey for SOFIA," P SP 110, pp. 339-364, 1998. 10.� W. G. Read, J. W. Waters, C. Smallcomb, E. Stone, Z. Shippony, D. L. Wu, A. Smedley, . Oltmans, D. Kley,

H. G. J. Smit, and J. L. Mergenthaler, "Microwave Limb Sounder UTH Validation," submitted to J. Geophys. Res. ,2000.

11.� H. J. Mastenbrook and R. E. Daniels, Atmospheric Water Vapor, ch. Measurements of stratospheric water vapor using a frost-point hygrometer. Academic Press, 1980.

12.� D. Rind, E. W. Chiou, W. Chu, S. Oltmans, J. Lerner, J. Larsen, M. P. McCormick, and L. McMaster, "Overview of the Stratospheric Aerosol and Gas Experiment II Water Vapor Observations: Method, Validation, and Data Characteristics," J. Geop'h s Res. 98, PR·,~8~l;;;lIi\l8~.....~~

13.� E. F. Erickson, M. R. Haas, S. W. J. Colgan, J. P. Simpson, and R. H. Rubin, "The Cryogenic Grating Spectrometer," in Airborne Astronomy Symposium on the galactic ecosystem: From gas to stars to dust, M. R. Haas, J. A. Davidson, and E. F. Erickson, eds., vol. 73 of ASP Conference Series, pp. 523-530, 1995.

14.� S. Colgan, NASA ARC, priv. comm., 1998. 15.� M. 1. Salby, Fundamentals of Atmospheric Physics, vol. 61 of Internat. Geophys. Series, Academic Press, 1996. 16.� J. W. Waters, W. G. Read, L. Froidevaux, R. F. Jarnot, R. E. Cofield, D. A. Fowler, G. K. Lau, H. M. Pickett,

M. L. Santee, D. L. Wu, M. A. Boyles, J. R. Burke, R. R. Lay, M. S. Loo, N. J. Livesey, T. A. Lungu, G. 1. Manney, 1. 1. Nakamura, V. S. Perun, B. P. Ridenoure, Z. Shippony, P. H. Siegel, and R. P. Thurstans, "The UARS and EOS Microwave Limb Sounder (MLS) Experiments," J.......A1.1!!!!.!JJh. Sci. 56, . 194-218, 1999.

To appear in Proc. SPIE 4014, 2000 11

MLS-derived zenith Water Vapor Overburden in precipitable microns 3-0CT-1992 FL370 3-0CT-1992 F'L390

80 80

70

60

."" 50 ~ j� 40

30

:l0

10 :l00 220 240 260

3-0CT-1992 FlAlO 80

70

80

50

40

30

:l0 20 4

10 280220 240 280 200 220 240 260 280

Ea.t Longitude East Longilude H,O 6: Mi••ing Oala - .: Scattering from Cloud. [lLm]

me: z __lLo_3 -OCT-1 992_UT...mod_CLAES.P' O:Odl;~~ ~b~~~'?ann::-~e-;:iv:it}~~~etJ~~h':-xll?~c~~~A

MLS-derived zenith Water Vapor Overburden in precipitable microns 4-0CT-1992 FL370 4-0CT-1992 F'L390

80 80

70

60

~ 50

:;ioS� 40

30

:l0

220 240 280

4-0CT-1992 FIA10 80

70

60

~ 50 3 ~ 40

30 30

:l0 4

10 220 240 260 280 200 220 240 260 280

Ea.t Longitude East. Longi l ude HID 6: MI••ing Data - .: Scattering from Clouds [lLm]

me: z.....--"_4-0CT-l 992_UT...mod_CLAES.p. 0: °d1<;~~ ~b~~~~ann~~e-;:i:~l}~~;;,leS1~~h':-xt?~c~P~A

Figure 9. Sample MLS WVO-maps faT October 3rd and 4tb 1992 as used for the statistical flightplanning analysis. Shown is the distribution of the water vapor over the continental US for four diH'erent flight levels, from 37,000 ft to 43,000 ft. The color code reHects the amount of precipitable zenith water vapor in microns.

To appear in Froc. SPIE 4014, 2000

20

12