Datasets:
ReadingTimeMachine
/
rtm-sgt-ocr-v1

Dataset card Data Studio Files Community
2
source
stringlengths
1
2.05k
⌀
target
stringlengths
1
11.7k
The historical light curve is shown in Fig.
The historical light curve is shown in Fig.
| and spans almost 40 years of observations.
\ref{0605_hist_lcurve} and spans almost 40 years of observations.
The longest light curve at an individual frequency covers more than 30 years at 8 GHz (see Table 1))
The longest light curve at an individual frequency covers more than 30 years at 8 GHz (see Table \ref{data}) ).
The individual light curves at 4.8. 8. 14.5. 22. and 37 GHz are shown in Fig. 2..
The individual light curves at 4.8, 8, 14.5, 22, and 37 GHz are shown in Fig. \ref{0605_lightcurve}.
The total-flux density variability of B0605—085 shows hints for à periodical pattern.
The total-flux density variability of $-$ 085 shows hints for a periodical pattern.
Four maxima have appeared at epochs around 1972.5. 1981.5. 1988.3. and 1995.8 with almost similar time separation.
Four maxima have appeared at epochs around 1972.5, 1981.5, 1988.3, and 1995.8 with almost similar time separation.
The last three peaks have similar brightness at centimeter wavelengths. whereas the first one is around | Jy brighter.
The last three peaks have similar brightness at centimeter wavelengths, whereas the first one is around 1 Jy brighter.
The data at 22 GHz and 37 GHz are less frequent and cover shorter time ranges of observations. but it is still possible to see that the last peak 1n 1995-1996 1s of similar shape and reaches the maximum at the same time as at lower frequencies.
The data at 22 GHz and 37 GHz are less frequent and cover shorter time ranges of observations, but it is still possible to see that the last peak in 1995-1996 is of similar shape and reaches the maximum at the same time as at lower frequencies.
The flares at 14.5 GHz show sub-structure with the dip in the middle of a flare of about 0.5 Jy. which ts ~30% of flare amplitude. suggesting more complex structure of the outburst or possible double-structure of the flares.
The flares at 14.5 GHz show sub-structure with the dip in the middle of a flare of about 0.5 Jy, which is $\sim$ of flare amplitude, suggesting more complex structure of the outburst or possible double-structure of the flares.
Figure 3 shows a close-up view of the complex sub-structure of flares at 14.5 GHz.
Figure \ref{0605_double} shows a close-up view of the complex sub-structure of flares at 14.5 GHz.
In order to check whether the flares might show the double-peak structure we have fitted the sub-structure of outbursts at 14.5 GHz with the Gaussian functions (dotted lines in Fig. 3).
In order to check whether the flares might show the double-peak structure we have fitted the sub-structure of outbursts at 14.5 GHz with the Gaussian functions (dotted lines in Fig. \ref{0605_double}) ).
The sum of the Gaussian functions. (solid line) follows the light curve fairly well and the possible double peaks have similar separation. they have time difference of 2.11 years between 1986.43 and 1988.54 sub-flares and 1.92 years between the 1994.45 and 1996.37 sub-flares.
The sum of the Gaussian functions (solid line) follows the light curve fairly well and the possible double peaks have similar separation, they have time difference of 2.11 years between 1986.43 and 1988.54 sub-flares and 1.92 years between the 1994.45 and 1996.37 sub-flares.
We searched also for archival data at optical wavelengths.
We searched also for archival data at optical wavelengths.
However. due to close proximity of a very bright foreground star of 9th magnitude. the optical observations of this quasar are difficult and there are not enough data for reconstruction of optical light curves.
However, due to close proximity of a very bright foreground star of 9th magnitude, the optical observations of this quasar are difficult and there are not enough data for reconstruction of optical light curves.
In order to check for possible periodical behavior of flares in the B0605—085 total flux-density radio light curves we have applied the Discrete Auto-Correlation Function (DACF) method (Edelson Krolik 1988. Hufnagel Bregman 1992) and the date-compensated discrete Fourier transform method (Ferraz-Mello 1981).
In order to check for possible periodical behavior of flares in the $-$ 085 total flux-density radio light curves we have applied the Discrete Auto-Correlation Function (DACF) method (Edelson Krolik 1988, Hufnagel Bregman 1992) and the date-compensated discrete Fourier transform method (Ferraz-Mello 1981).
The discrete auto-correlation function method permits to study the level of auto-correlation in unevenly sampled data sets avoiding interpolation or addition of artificial data points.
The discrete auto-correlation function method permits to study the level of auto-correlation in unevenly sampled data sets avoiding interpolation or addition of artificial data points.
The values are combined in. pairs (uj.bj). for each O<ij€N. where N is the number of data points.
The values are combined in pairs $(a_{i}, b_{j})$, for each $0 \leq i,j\leq N$, where N is the number of data points.
First. the unbinned discrete correlation function is calculated for each pair where @ b are the mean values of the data series. and σι. Tp are the corresponding standard deviations.
First, the unbinned discrete correlation function is calculated for each pair where $\bar{a}$, $\bar{b}$ are the mean values of the data series, and $\sigma_{a}$, $\sigma_{b}$ are the corresponding standard deviations.
The discrete correlation function values (DCF) for each time range Ar;=fj—t; are calculated as an average of all UDCF values. which time interval fall into the range c—Ar/2€An;t+Δτ/2. where r is the center of the bin.
The discrete correlation function values (DCF) for each time range $\Delta t_{ij} = t_{j} - t_{i}$ are calculated as an average of all UDCF values, which time interval fall into the range $\tau - \Delta \tau / 2 \leq \Delta t_{ij} \leq \tau + \Delta \tau / 2$, where $\tau$ is the center of the bin.
The higher is the value of Ar the better is the accuracy and the worse is time resolution of the correlation curve.
The higher is the value of $\Delta \tau$ the better is the accuracy and the worse is time resolution of the correlation curve.
In case of the auto-correlation function. the signal is cross-correlated with itself and a=5. à=b. and Cu=op.
In case of the auto-correlation function, the signal is cross-correlated with itself and $a=b$, $\bar{a}=\bar{b}$, and $\sigma_{a}=\sigma_{b}$.
The error of the DACF is calculated as a standard deviation of the DCF value from the group of unbinned UDCF values The DACF method yields different numbers of UDCF per bin. which can affect the final correlation curve.
The error of the DACF is calculated as a standard deviation of the DCF value from the group of unbinned UDCF values The DACF method yields different numbers of UDCF per bin, which can affect the final correlation curve.
In order to check the results of the DACF we also applied the Fisher z-transformed DACF method. which allows to create data bins with equal number of pairs (Alexander 1997).
In order to check the results of the DACF we also applied the Fisher z-transformed DACF method, which allows to create data bins with equal number of pairs (Alexander 1997).
The Date-Compensated Discrete Fourier Transform (DCDFT) method was created in order to avoid the problem of finiteness of the Fourier transform and allows us to estimate timescales of variability in unevenly sampled data with better precision.
The Date-Compensated Discrete Fourier Transform (DCDFT) method was created in order to avoid the problem of finiteness of the Fourier transform and allows us to estimate timescales of variability in unevenly sampled data with better precision.
The method is based on power spectrum estimation fitting sinusoids with various trial frequencies to the data set.
The method is based on power spectrum estimation fitting sinusoids with various trial frequencies to the data set.
In case of a periodic process formed by several waves. the DCDFT method permits to filter the time series with an already known frequency and find additional harmonies.
In case of a periodic process formed by several waves, the DCDFT method permits to filter the time series with an already known frequency and find additional harmonics.
We searched for periods in the light curves at 4.8 GHz. 8 GHz. and 14.5 GHz. which span more than 30 years.
We searched for periods in the light curves at 4.8 GHz, 8 GHz, and 14.5 GHz, which span more than 30 years.
We have removed a trend before applying a periodicity analysis.
We have removed a trend before applying a periodicity analysis.
The trend was estimated by fitting a linear regression funetion into the light curve.
The trend was estimated by fitting a linear regression function into the light curve.
The auto-correlation method reveals periods of 7,740.2 yr at 14.5 GHz. 7.240.1 yr at 8 GHz. and 8.040.2yr at 4.8 GHz with high correlation coefficients of 0.84. 0.77. and 0.83 respectively.
The auto-correlation method reveals periods of $\pm$ 0.2 yr at 14.5 GHz, $\pm$ 0.1 yr at 8 GHz, and $\pm$ 0.2yr at 4.8 GHz with high correlation coefficients of 0.84, 0.77, and 0.83 respectively.
We have estimated the periods by fitting a Gaussian function. into the DACF peak.
We have estimated the periods by fitting a Gaussian function into the DACF peak.
The error. bars are calculated as the error bars from the fit.
The error bars are calculated as the error bars from the fit.
The Fisher z- method for calculating DACF provides the same results.
The Fisher z-transformed method for calculating DACF provides the same results.
An example of calculated discrete auto-correlation
An example of calculated discrete auto-correlation
|ue and FIR.
blue and FIR.
This is because we have the freedom. au indeed: the motivation from the ποσο], to cumulate the ellects of collisions.
This is because we have the freedom, and indeed the motivation from the model, to cumulate the effects of collisions.
In contrast. semi-analvtic hierarchica models form spheroids over a Hubble time at the redshifi at which galaxy mergers occur.
In contrast, semi-analytic hierarchical models form spheroids over a Hubble time at the redshift at which galaxy mergers occur.
This may not be enough if the dust is dispersed by the ensuing starburst.
This may not be enough if the dust is dispersed by the ensuing starburst.
Introduction ofan delay between multiple starburst trigerrecl by different orbital stages during the pre-merecr phase mieh solve this.
Introduction of an delay between multiple starburst trigerred by different orbital stages during the pre-merger phase might solve this.
Alternatively one might appeal to adcditiona sources of Ht luminosity such as the one that might be associated with AGN formation.
Alternatively one might appeal to additional sources of IR luminosity such as the one that might be associated with AGN formation.
All indications however suggest that AGNs do not play a major role in accounting or most of the observed. submim sources. as evidence rom theCHANDRA deep fields.
All indications however suggest that AGNs do not play a major role in accounting for most of the observed submm sources, as evidenced from the deep fields.
In our case we have a yhvsical model that works in the right direction and. vields a sel-consistent panchromatic explanation of the observec universe both locally and at. high. redshift.
In our case we have a physical model that works in the right direction and yields a self-consistent panchromatic explanation of the observed universe both locally and at high redshift.
We are aware hat what we called spheroids at high redshifts but. are in he process of a violent burst of star formation will no nave the morphologies of a classical carly type galaxy. as it will take some time for these objects to relax.
We are aware that what we called spheroids at high redshifts but are in the process of a violent burst of star formation will not have the morphologies of a classical early type galaxy, as it will take some time for these objects to relax.
We defer a more detailed study of the evolution of relaxed galaxies to a companion paper. but note that it is likely that this fraction will be fairly small at zc3 since there will be very Little time elapsed since the collision.
We defer a more detailed study of the evolution of relaxed galaxies to a companion paper, but note that it is likely that this fraction will be fairly small at $z \geq 3$ since there will be very little time elapsed since the collision.
llowever. in spite of this. issue. an important consequence of our model. is that the FER counts and Iuminosity density are expected to continue to rise. with redshift to z~3. Hence we predict that a future mission with adequate resolution to avoid. source confusion. which sadly may not be the case for orSIRTLE. will provide an important test of the galaxy collision model.
However, in spite of this issue, an important consequence of our model, is that the FIR counts and luminosity density are expected to continue to rise with redshift to $z\sim 3.$ Hence we predict that a future mission with adequate resolution to avoid source confusion, which sadly may not be the case for or, will provide an important test of the galaxy collision model.
We thank Cuilaine Lagache. Hervé Dole and. Jean-Loup Puget for providing us with their results before publication.
We thank Guilaine Lagache, Hervé Dole and Jean-Loup Puget for providing us with their results before publication.
We also acknowledge a constructive referee. report. from. which the presentation of the model has benefited a lot.
We also acknowledge a constructive referee report from which the presentation of the model has benefited a lot.
J.D. is supported by the Leverhulme trust.
J.D. is supported by the Leverhulme trust.
Figure 9 shows the same results presented in Table 2 but plotted as the ratio to the fiducial case.
Figure \ref{fig:ratio_modeluncertainty} shows the same results presented in Table \ref{table:GRBrate_theory2} but plotted as the ratio to the fiducial case.
The red solid and blue dot-dashed lines represent the result forPdet andAdet, respectively.
The red solid and blue dot-dashed lines represent the result for and, respectively.
'The detection rate for is about twice as large as the fiducial case.
The detection rate for is about twice as large as the fiducial case.
This fact remains unchanged for different Taelay, Which is represented in the top panel of Figure 6 with the light-blue dashed curve.
This fact remains unchanged for different $\tau_{\rm delay}$, which is represented in the top panel of Figure \ref{fig:delay-vs-PAdet} with the light-blue dashed curve.
This is because Rextra of 0.1 makes the bursts with intrinsically soft Band spectrum with BS—2.5 detectable; these soft bursts account for about half of Alert andC'TAobs samples.
This is because $R_{\rm extra}$ of 0.1 makes the bursts with intrinsically soft Band spectrum with $\beta \lesssim -2.5$ detectable; these soft bursts account for about half of and samples.
By contrast, Rextra=0.1 has little influence on the photon-count distribution and the redshift distribution.
By contrast, $R_{\rm extra}=0.1$ has little influence on the photon-count distribution and the redshift distribution.
Comparing2 with3, the latter shows a larger increase of the detection rate compared to the former.
Comparing with, the latter shows a larger increase of the detection rate compared to the former.
This simply reflects that the fraction of bursts with 8>—2 is less than those with 8x;—2.5.
This simply reflects that the fraction of bursts with $\beta>-2$ is less than those with $\beta \lesssim -2.5$.
We also calculate the expected LAT detection rate of 2 and3, and then obtain 24 yr! and 14 yr!, respectively.
We also calculate the expected LAT detection rate of and, and then obtain 24 $^{-1}$ and 14 $^{-1}$, respectively.
The former is more than three times as large as the observed LAT rate for the bursts of Ton> 2sec and z«5 (about 7-8 yr~'), which seems unrealistic.
The former is more than three times as large as the observed LAT rate for the bursts of $T_{90}>2$ sec and $z<5$ (about 7–8 $^{-1}$ ), which seems unrealistic.
The calculated LAT event rate for our fiducial case and3 are similar to each other and about 1.5-2.0 times as large as the observed one.
The calculated LAT event rate for our fiducial case and are similar to each other and about 1.5–2.0 times as large as the observed one.
These somewhat large rates appear to be consistent with the analysis by Beniaminietal.(2011),, claiming the existence of the spectral softening below the LAT band in some of bright bursts, since we have not taken into account anyspectral softening feature in our simulations.
These somewhat large rates appear to be consistent with the analysis by \citet{Beniamini2011}, claiming the existence of the spectral softening below the LAT band in some of bright bursts, since we have not taken into account anyspectral softening feature in our simulations.
We discuss this point a bit more in Section 5..
We discuss this point a bit more in Section \ref{sec:summary}.
In addition to the cases summarized in Table 2,, we studied the dependence of the detection rate on the luminosity function (see Eq. (2))
In addition to the cases summarized in Table \ref{table:GRBrate_theory2}, , we studied the dependence of the detection rate on the luminosity function (see Eq. \ref{eq:rho_z}) )
and (3)), the normalization of the effective area, fA (see Section 3.3)), and the low-energy end of the LSTs sensitivity, Flow (see Eq. 10)).
and \ref{eq:phi_L}) )),the normalization of the effective area, $f_A$ (see Section \ref{subsec:LSTs}) ), and the low-energy end of the LSTs sensitivity, $E_{\rm low}$ (see Eq. \ref{eq:Elow}) ).
For the luminosity function, we found that when one of the 6 parameters included is varied from the best fit value within its errors, the rate changes at most by a factor of ~ 0.8-1.3 for bothPdet andAdet.
For the luminosity function, we found that when one of the 6 parameters included is varied from the best fit value within its errors, the rate changes at most by a factor of $\simeq$ 0.8–1.3 for both and.
For fa, we see its influence on the detection rate by varying the value from 0.3 to 3.
For $f_A$, we see its influence on the detection rate by varying the value from $0.3$ to $3$.
At this time, as a function of fa, the background count rate Rpg shifts from about 0.03 Hz to 3 Hz to keep the sensitivity of our LSTs model consistent with the official one.
At this time, as a function of $f_A$, the background count rate $R_{\rm bg}$ shifts from about $0.03$ Hz to $3$ Hz to keep the sensitivity of our LSTs model consistent with the official one.
We found that this range of fa changes the detection rate by a factor of 0.8-1 forPdet, though the change of the rate is negligible.
We found that this range of $f_A$ changes the detection rate by a factor of 0.8–1 for, though the change of the rate is negligible.
This reflects the fact that for non-detected prompt emissions, the detection conditions (1) and (2) (which are described in the last part of Section 3.2)), have comparable importance, while for non-detected afterglows, the condition (2) is the most strict.
This reflects the fact that for non-detected prompt emissions, the detection conditions (1) and (2) (which are described in the last part of Section \ref{subsec:detect_conditionCTA}) ), have comparable importance, while for non-detected afterglows, the condition (2) is the most strict.
Note that the expected photon counts ΛΙ can vary in proportion to f4.
Note that the expected photon counts $N_\gamma$ can vary in proportion to $f_A$.
For Ew, we see the detection rate for the variousexponent of cosine in Eq.(10)).
For $E_{\rm low}$, we see the detection rate for the variousexponent of cosine in\ref{eq:Elow}) ).
By varying it from —4.3 to 0, we found that the detection rate changes by a factor of ~ 1.0-1.2 for Pdet and ~ 0.9-1.1for Adet.
By varying it from $-4.3$ to 0, we found that the detection rate changes by a factor of $\simeq$ 1.0–1.2 for and $\simeq$ 0.9–1.1for .
Therefore the luminosity
Therefore the luminosity
inulti-lobed. n:vure of the coutrition fuwction in Fig.δν,
multi-lobed nature of the contribution function in Fig.\ref{cf},
which shows sensitivity throtehout the 1-10X μα range.
which shows sensitivity throughout the 1-1000 $\mu$ bar range.
We attempt to preserve the smooth lature of T(p) inferred frou ISO spectra).. but we οaution hat t1e rescence of rea oscillations in the eniperatiro profile (due ο stratospheric wave activity. for exanupl|) would uot b detected via tjs supο ραιοcrisatiou.
We attempt to preserve the smooth nature of $T(p)$ inferred from ISO spectra, but we caution that the presence of real oscillations in the temperature profile (due to stratospheric wave activity, for example) would not be detected via this simple parameterisation.
At the same nue. we scale the abundance above the tro?opatSC and minimised the fit to the data. \7. over the 7.1-8.3 jan (1200-1LOO) 1 ) specral range.
At the same time, we scaled the $_4$ abundance above the tropopause and minimised the fit to the data, $\chi^2$, over the 7.1-8.3 $\mu$ m (1200-1400 $^{-1}$ ) spectral range.
The four temiperatire xofile parameerisations were as follows: Au cxalpe of. the «D analysisH dis- she»vu in Fie. l..
The four temperature profile parameterisations were as follows; An example of the $\chi^2$ analysis is shown in Fig. \ref{Tch4},
which shows a diueusioual 47. surface) in twothe ceutrc. fect the spectral efof varving the CIT; mole fraction ou the left and the temperature profile ou the right.
which shows a two dimensional $\chi^2$ surface in the centre, the spectral effect of varying the $_4$ mole fraction on the left and the temperature profile on the right.
The two correlates paralneters can be mareially separated because of the different 1uorphologies of their effects on the spectrum.
The two correlated parameters can be marginally separated because of the different morphologies of their effects on the spectrum.
The best-fitting teuiperature structure axd the error range (taking the auticorrelation with CII, oeito. account) are shown in Fig. 5..
The best-fitting temperature structure and the error range (taking the anticorrelation with $_4$ into account) are shown in Fig. \ref{temp},
conrpared to T(p) cternunations from. previous stidies.
compared to $T(p)$ determinations from previous studies.
The WOOl axd COO extrema of the error range were used in the error analvsis for composijon in the followiug sections.
The warm and cool extrema of the error range were used in the error analysis for composition in the following sections.
The best-fitting temiperatire profile provides an optima CIT, luoe fraction of (9.0430)«10.Lat 50 mbar. decreasing to {r9£03)«101 at 1l pbar.
The best-fitting temperature profile provides an optimum $_4$ mole fraction of $(9.0\pm3.0)\times10^{-4}$ at 50 mbar, decreasing to $(0.9\pm0.3)\times10^{-4}$ at 1 $\mu$ bar.
However. using the cool au worn extrema of the temperature profile. the CIT, luoe fraction could vary between 6.0js.0«10.4,
However, using the cool and warm extrema of the temperature profile, the $_4$ mole fraction could vary between $6.0-18.0\times10^{-4}$.
The telorature and CIT, determunations will be cisceussed iu Section 5..
The temperature and $_4$ determinations will be discussed in Section \ref{discuss}.
Fie.
Fig.
G shows the effect on the variation of the D/IILratio iu CTL, ou the AKARI spectrum.
\ref{ch3d} shows the effect on the variation of the D/H ratio in $_4$ on the AKARI spectrum.
The 1g vibrational uid of CUD at 1161 cin! is seen eutizelv in euiission. affecting the shape o‘the spectrum on the short-wavelength side of the SC2 spectitm.
The $\nu_6$ vibrational band of $_3$ D at 1161 $^{-1}$ is seen entirely in emission, affecting the shape of the spectrum on the short-wavelength side of the SG2 spectrum.
Unfortunately. the absolute caliration (both in terusof flux. aud the wavelength calibration) has the largest ¢eeree of uncertainty at the cud of the crannel: spectral reductions optimized for the two promineit aicl-IR eatures. CoG and CIT,. provided differing esnates oft1ο flux di this region. so the optimization for CIT, haea to be used.
Unfortunately, the absolute calibration (both in termsof flux, and the wavelength calibration) has the largest degree of uncertainty at the end of the channel: spectral reductions optimized for the two prominent mid-IR features, $_2$ $_6$ and $_4$, provided differing estimates of the flux in this region, so the optimization for $_4$ had to be used.
Furthermore. as no spectral lijos are resolved. this can pxwide no more than ai order of magnitude estinate of he CII4D abundance.
Furthermore, as no spectral lines are resolved, this can provide no more than an order of magnitude estimate of the $_3$ D abundance.
Nevertjcless. we scaled the CIT, profile bv a CTI;D/CTL, ratio beween 2.0410! and 6.0.10! in steps of 0.1«10. using the cool wari and best-fitting T(p) structures frou the previous section.
Nevertheless, we scaled the $_4$ profile by a $_3$ $_4$ ratio between $2.0\times10^{-4}$ and $6.0\times10^{-4}$ in steps of $0.1\times10^{-4}$, using the cool, warm and best-fitting $T(p)$ structures from the previous section.
Minimisine 4? over the 1100-12σι t spectral range. we find that D/II ratios 11i CIT, in he raree COdE1.0)«10! can fit the spectrua. CDIESent with the value of 3.35«10 that we would expect from the ISO-denrved D/II ratio iu Il» of 6.5&10D(?).. using a fractionation factor of f=1.25 betwee1 deuterini iu Πο aud CIT, (?).
Minimising $\chi^2$ over the 1100-1200 $^{-1}$ spectral range, we find that D/H ratios in $_4$ in the range $(3.0\pm1.0)\times10^{-4}$ can fit the spectrum, consistent with the value of $3.25\times10^{-4}$ that we would expect from the ISO-derived D/H ratio in $_2$ of $6.5\times10^{-5}$, using a fractionation factor of $f=1.25$ between deuterium in $_2$ and $_4$ .
. This D/II ratio in CIT, is pertiucu to the 0.2 mbar level. the peak of the CIT4D coitributiou function in Fie.
This D/H ratio in $_4$ is pertinent to the 0.2 mbar level, the peak of the $_3$ D contribution function in Fig.
2 The second mios prominent feaure of the mid-IR AIKARI spectrun is the ethane peak a 12.2 gau. which is fitted in Fie. 7..
\ref{cf} The second most prominent feature of the mid-IR AKARI spectrum is the ethane peak at 12.2 $\mu$ m, which is fitted in Fig. \ref{c2h6}.
Usine the three temerature profiles. we scale the Coll; profile iu steps of LI between (LO and 2.0. to derive a best fit of US+0.2 times f1ο profile.
Using the three temperature profiles, we scale the $_2$ $_6$ profile in steps of 0.1 between 0.0 and 2.0, to derive a best fit of $0.8\pm0.2$ times the profile.
The error rauge was calculated using the three models represeutiug the ceure auc the two extrema of the temperature profiles. vieldiug 0.6 for the waruest profile. 1.0 for the coldest profile. aud L8 as the optinuun profile.
The error range was calculated using the three models representing the centre and the two extrema of the temperature profiles, yielding 0.6 for the warmest profile, 1.0 for the coldest profile, and 0.8 as the optimum profile.
This vields à απ! abuidance of (2.540.6)ς1089 at 3.1 prbar. iux (AHA?A10 “at the peas of the ethane coutributiou function at 0.3 nixw in Fig. 2..
This yields a maximum abundance of $(2.5\pm0.6)\times10^{-6}$ at 3.4 $\mu$ bar, and $(8.5\pm2.1)\times10^{-7}$ at the peak of the ethane contribution function at 0.3 mbar in Fig. \ref{cf}. .
The longavaveleneth exd of the SC2 spectrum is sensitive to he abundance of acetylene (ΟΠ), where we see some discrepancy beweenthe model and the data in Fig. 7..
The long-wavelength end of the SG2 spectrum is sensitive to the abundance of acetylene $_2$ $_2$ ), where we see some discrepancy betweenthe model and the data in Fig. \ref{c2h6}.
noted the cifficultics associated with simultarwcously modeling the abundances of Cog
noted the difficulties associated with simultaneously modelling the abundances of $_2$ $_6$
The final catalogues were created by matching the detections within the single colour catalogues to produce one master catalogue for each field, containing magnitudes and errors in all bands, and the corresponding extraction flags.
The final catalogues were created by matching the detections within the single colour catalogues to produce one master catalogue for each field, containing magnitudes and errors in all bands, and the corresponding extraction flags.