Interactive comment on “ Reduction in Earth Reflected Radiance during the Eclipse of 21 August 2017 ”

The paper has considerably improved. Most of the my comments and recommendations have been already mentioned by reviewer #1. From my point of view the Maignan et al., 2004 study has to be reported in the introduction sections and the "differences" observed with this study have to be reported in more detail in the section that currently this work is mentioned. I think this is an interesting and unique work that should be published in AMT.

1.In the abstract and also later in the text (l.237 ff, l. 366 ff), the observations are compared to modelling results by Emde and Mayer, 2007: "A previously published clear-sky model (Emde and Mayer, 2007) shows results for a nearly overhead eclipse had R EN (340nm)=1.7x10ˆ4compared to the maximum measured non-averaged R EN (340) at Casper of 515±27 with optically thin clouds under similar geometrical conditions."Such a quantitative comparison is not possible, because the modelling result refers to the reduction of global irradiance (measured from the surface) in the center of the umbral shadow.This is a different quantity and the observation geometry is completely different, thus it is not surprising that the results do not agree to the DISCOVR observations.For a quantitative comparison, the 3D radiative transfer model needs a completely different setup, one has to model the reflected radiance for the specific observation geometry of DISCOVR (with a phase angle of about 172 • ).This should be possible using a Monte Carlo code like MYSTIC used by Emde and Mayer, 2007, but as said before, it requires a completely different setup.It can be mentioned in the text, that it would be interesting to model the observations with 3D RT models, but quantitative comparisons should be removed from abstract, text and summary.
2. Why is the global reduction of the reflectance (Section 3.2) only calculated for Casper and not for Columbia.It would be interesting to see, how much the reduction of global reflectance depends on the location of the clouds.Please include results for Columbia in Section 3.2.
3. Could the DISCOVR data, which is used for the study, added as supplementary data to the paper in addition to the provided links?C2

Interactive comment
Printer-friendly version Discussion paper Specific comments: l.1 "Sunlit" -> "sunlit" l. 16: "A reduction of 9.7±1.7% in the radiance (387 to 781 nm) reflected from the Earth towards L1 was obtained..." -> Please clarify that this is the spectrally integrated global reflectance.l. 25ff:"A previously published ..." -> remove this sentence from abstract (see above).l.45 "earth" -> "Earth" l. 55: "The totality region (umbra) is about 250 to 267 km in diameter, ..." -> in line 151 it is said that the totality shadow is 116 km wide over Caspar.How do these numbers match?In l. 151ff both axes of the umbral shadow should be provided for both locations (Caspar and Columbia).On which parameters does the size of the shadow depend apart from solar zenith angle?The distances Sun-Earth, Moon-Earth also determine the extent of the shadow.l.119: "To reduce the volume of data, all measurements, except those from the 443 nm channel, were averaged onboard DSCOVR to 1024 x 1024 pixels."-> Why is the 443nm channel treated differently?l. 167: "340 nm, with strong Rayleigh scattering effects (haze)" -> haze (aerosol) scattering is not the same as Rayleigh (molecular) scattering.2.  l. 237ff: see "general comment 1" Fig. 8 a,b,c, A2: I think that not all of these figures are needed.Fig. 3 nicely shows, how the shadow and the Earth looks in various channels and Fig. 4 shows the synoptical conditions during the eclipse and the day before and after the eclipse.I suggest to put most of these figures in the appendix.l. 285: "While the figures are similar from wavelength to wavelength, there are differences in the depth of the eclipse totality and the reflectivities of the surrounding clouds."With the chosen grey-scale colormaps these difference are not visible.I suggest to use a colormap similar to the one in Fig. 7 to visualize the differences in the depth of the eclipse totality.l. 325: "This means that EPiC is observing close to "hotspot" conditions where the backscatter amount increases with increasing wavelength (Maignan et al., 2004).At 551 and 680 nm the hotspot effect is smaller than at 780 nm." -> This is not obvious from Table 3, integrated counts are much larger for 551nm than for 780nm.Please explain.
l. 330: "The solar spectrum used is a combination of data named atlas_plus_modtran (Mayer and Kylling, 2005)."-> Mayer and Kylling is the reference for the libRadtran software package, from which the solar irradiance data is taken.Please rewrite sentence more clearly.
l. 60ff: The overview paper ofGerasopoulos et al., 2008  is cited and it is said that it would include MODIS observations of the eclipse from 2006 over Europe.This is not correct, the paper includes a MODIS image from the same day (taken before eclipse) to show the cloud formation over Greece.l.66ff: "A 3D Monte Carlo radiative transfer study (Emde and Mayer, 2007) was applied to the geometry for the nearly overhead total eclipse of 29 March 2006 (13:20 local time in Turkey), but without the effect of clouds included in the calculation.Successful modelling of an eclipse under realistic conditions is the first step to improved modelling ..." -> The modelling was realistic for the given observation over Greece, because the region was cloud-free.A comparison to observations showed an excellent agreement (see Kadzantzidis et al. 2007, https://www.atmos-chem-phys.net/7/5959/2007/).Of course, clouds need to be included when present.Please clarify this sentence.l.71ff: "The observations from the DSCOVR satellite are part of a larger project that C3 satellite and ground-based measurements using a pyranometer(Ji and Tsay, 2000)  and the Pandora Spectrometer Instrument(Herman  et al., 2009)  at both sites."-> Is this data available?l. 75: "This study presents the only synoptic satellite data of the sunlit Earth ever obtained during an eclipse ..." -> What about images from geostationary satellites?I remember movies of MSG images of the eclipse from March 2006.

Fig. 3 :
Fig.3: These images are very nice.I would prefer north up as usual, even if inconsistent with Fig. 7. l. 173 "3.1 Comparison of Eclipse and Non-Eclipse Days for Caspar, WY and Columbia, MO "-> Use a more specific title, what is compared?Fig. 5: "Middle"-> "Bottom" Fig. 6a,b: These figures all look quite similar.Why are all channels shown for Caspar, WY and only one for Columbia (as lower left plot in Fig 6a).This arrangement is confusing.Suggestion: Use 3 representative channels and show the results for Caspar, WY on the left and the corresponding results for Columbia on the right.Maximum values for all channels and for both locations should be included in Table2.
l. 234ff: "A detailed radiative transfer study for realistic conditions is made feasible by using EPIC's simultaneous estimates of cloud reflectivity and transmission, cloud height, aerosol amounts, and ozone amounts."-> It there a data product including C4 yes, please provide reference.
l. 363: "A previously published clear-sky model result for a nearly overhead eclipse ratios and an ocean surface albedo of 0.06 (Emde and Mayer, 2007) had R_EN (340nm)=1.7x10ˆ4compared to the measured non-averaged R_EN (340) at Casper of 515±27 with optically thin clouds under similar geometrical conditions."-> these results can not be compared (see general comments).Not R_EN has been modelled by Emde and Mayer, the value refers to global irradiance at the surface!C5

Table 2 :
Please include also the <R_EN> values for Columbia in the table.

Table 3 :
Please include this table also for Columbia.