Utilization of the Japanese meteorological satellite Himawari-8 as an infrared space telescope for lunar and planetary science

Introduction

Since 2015, the Japanese meteorological satellite Himawari-8 has been observing weather in the Asia-Pacific region every 10 minutes [1]. Scanning the full-disk Earth also covers space adjacent to the Earth, where celestial objects, such as the Moon, are occasionally included (Figure 1). Onboard Himawari-8 is an Advanced Himawari Imager (AHI), whose high spatio-temporal resolution can provide over 900 lunar images with 23 km/pix or better.

Furthermore, various AHI bands from visible to infrared wavelengths possibly provide multispectral information about planets. In particular, some of the nine infrared wavelength bands on AHI have not been used for observing the Moon by any other spaceborne telescopes. For example, within the AHI infrared coverage of 6–14 μm, the Diviner radiometer onboard Lunar Reconnaissance Orbiter (LRO) has only three channels around 8 μm [2].

Despite such potential, the AHI images have never been recognized as a useful dataset in planetary science. In order to demonstrate the possibility of utilization of AHI data, we investigate the lunar thermal environment with the AHI data. We develop the procedure to extract lunar multiband brightness temperatures. Then, we compare them with the Diviner measurements and numerical simulations, posing some constraints on lunar thermophysical conditions.

Figure 1. Synthesized images of the Moon captured within AHI.

 

Method

To derive lunar brightness temperatures, we use the Himawari Standard Data (HSD) published by the Japanese Meteorological Agency. First, we extract lunar images from all HSDs taken by the end of November 2021. Next, we manually choose 248 lunar images in which more than a half part of the Moon is captured (Figure 2). Then, we calculate the lunar orientation with SPICE to obtain the lunar longitude and latitude of each HSD pixel. Finally, after estimating background noise, we convert the radiance value to brightness temperatures at all nine bands.

To interpret the derived brightness temperatures, we employ a 1-D conductive model to simulate the temperature on the Moon. In this simulation, the material differences must be considered because the thermal inertia of the media influences the temperature prediction, particularly on the nightside. Thus, the calculation is conducted for both rock and regolith, using the Apollo data [3]. Moreover, surficial roughness is critical because of its effects on shadows, notably in the morning and evening. Thus, we incorporate the statistical model among incidence angle, emission angle, shadow ratio, and roughness parameterized by the RMS mean slope [4, 5].

Figure 2. (a, b) Images of the half-Moon and almost-new Moon taken on 2018/12/28 08:20 and 2015/11/09 13:10 in UTC, respectively.

 

Brightness temperatures

Lunar brightness temperatures derived from HSD are consistent with those measured by Diviner. Due to the similarity between band 11 on AHI (8.40–8.78 μm) and channel 5 on Diviner (8.38–8.68 μm), we compare HSD with the Diviner Global Cumulative Product (GCP) [6]. Although HSD is saturated at local times from 8 to 16 hours due to its dynamic range, the brightness temperatures from HDS match those from Diviner GCP in the morning and evening. In addition, HSD at Tycho crater also agrees with Diviner GCP even in the nightside, indicating the reliability of AHI for planetary science.

Multiband lunar images by AHI reveal differences in brightness temperature among wavelengths, so-called anisothermality. In the morning and evening, the brightness temperature differences among bands increase with the incident angle. Nightside temperature differences also exist in Tycho crater. These anisothermal features are consistent with the Diviner measurements and indicate the mixture of various temperatures even within a pixel of HSD.

 

Lunar thermophysical conditions

The anisothermality in the morning and evening is caused by the surface roughness of the Moon. Due to various slopes on millimetric to centimetric scales, the surface temperature on the Moon shows a wide-ranging distribution, increasing anisothermality with the incident angles. In addition, our radiance simulation with the RMS mean slope of 16–20 degrees is consistent with the observed values at incident angles lower than 70 degrees. These slope angles also match those at Apollo landing sites [7].

On the other hand, the nightside anisothermality reflects rock abundance. Because of the high thermal inertia of rocks, the midnight temperature of rocks is expected to be higher than that of regolith by over 100 K. As a result, only a slight mixture of rock in regolith can make brightness temperatures higher than regolith (Figure 3). Our mixture model of rock and regolith estimates the rock abundances at the equator and Tycho crater are 0.18–0.48 and 6.1–10.3%, respectively. These estimates are also consistent with the Diviner constraints [8].

Figure 3 Temporal temperature variations at the lunar equator after sunset. The black points are the HSD brightness temperatures at band 14. The black dashed and dotted lines show the rock and regolith temperatures in our simulation. The solid lines show rock-regolith mixture models with rock ratios of 0.2, 0.5, 2, 10, 25, and 50 %.

 

Conclusion

Consistencies between AHI and Diviner indicate that the quality of HSD is sufficiently high for planetary science. Despite the spatial resolution lower than Diviner, the unique wavelength coverages of AHI can provide new spaceborne infrared datasets. In addition, AHI sometimes captures other planets like Mercury, Venus, Mars, and Jupiter. Thus, HSD can be a valuable data source for calibrations and future planetary sciences.

 

Note

The contents of this presentation have been submitted by Nishiyama et al. to Earth, Planets, and Space.

 

Reference

[1] Okuyama, A. et al., 2018, Journal of the Meteorological Society of Japan 96B, 91.

[2] Paige, D.A. et al., 2010, Space Science Reviews 150, 125–160.

[3] Horai, K. & Simmons, G., 1972, Thermal Characteristics of the Moon. pp. 243–267.

[4] Bandfield, J.L. et al., 2015, Icarus 248, 357–372.

[5] Davidsson, B.J.R. et al., 2015, Icarus 252, 1–21.

[6] Williams, J.P. et al., 2017, Icarus 283, 300–325.

[7] Helfenstein, P. & Shepard, M.K., 1999, Icarus 141, 107–131.

[8] Bandfield, J.L. et al., 2011, JGR: Planets 116.