Validation of estimated cloud fraction from MERRA-2 and AIRS data using ground based observation over IAO, Hanle

Highlights

• Number of clear nights are quantified using cloud fraction (CF) data at various astronomical sites.

• Observed CF data from visual observation are validated with MERRA-2 and AIRS data at IAO-Hanle

• Visual CF data shows good correlation of 0.78–0.84 with AIRS (ascending) data at the site

• Median CF for night at Hanle and Mauna Loa are 0.32 and 0.49,respectively using MERRA-2.

Abstract

We report for validation of cloud fraction estimated from high-resolution (hourly) data from MERRA-2 and AIRS instrument aboard NASA’s Aqua satellite during its ascending and descending orbits with hourly ground-based visual observation taken at Indian Astronomical Observatory (IAO), Hanle during 1997–2018. The study reveals 70% useful nights, equivalent to 254 spectroscopic nights in a year. A comparative study of cloud fraction was performed using homogeneous and continuous data spanning six years (2013–2018) from MERRA-2 and AIRS over IAO, Hanle and Merak sites in the Himalayan region, and several astronomical observatories located in Atacama desert in Chile, Canary islands in Spain, National Astronomical Observatory (NAO), Mexico, and Mauna Loa in Hawaii. The study reveals cloud fractions of 0.07–0.11 at 25% (1st Quartile) and 0.32–0.40 at 50% (median) at the Himalayan sites. On the other hand, the median cloud fractions are 0.06–0.13 at the Chilean sites, 0.16 at NAO, and 0.29 at Canary islands. The observed 25% of the cloud fraction at the Himalayan sites during the lowest three cloud-fractional months (October-December) are comparable with other sites located in Chile and Canary islands. The median cloud fraction at Hanle is slightly lower than Mauna Loa. The study further concluded that the percentage of useful nights obtained from the visual observation (70%) at IAO-Hanle is very close to the percentage of clear sky (68%) estimated from the MERRA-2 data.

Introduction

Cloud coverage is one of the most important parameters to characterize or assess the quality of ground based astronomical observatories operating in optical and near-infrared spectral region, whereas there is little importance of cloud coverage for astronomical observatories operating in radio or low-frequency region. About 65% of the Earth’s surface is covered by clouds. Due to various thermodynamic processes, clouds exhibit large dynamical variation on diverse temporal and spatial scales, which make it difficult to quantify or characterize cloudiness over short temporal timescales. There are various methods to observe cloud fraction or cloud coverage from surface, in situ (aircraft) or from space based techniques (Wielicki et al., 1996, Sassen and Wang, 2008). Each technique has its advantages and disadvantages in different applications. The practice of continuous visual observation is possible for the observatories which have sufficient local staff. On the other hand, automated all-sky CCD or Infrared (IR) cameras may be employed to monitor cloud coverage (Long et al., 2006, Feister et al., 2010, Sharma et al., 2015). However, estimation of cloud coverage from such instruments is affected by contrast of sky brightness emission during different phases of moon, local light pollution, and other atmospheric dynamics etc. (Shamir and Nemiroff, 2005). Further, such data may not be archived or publicly available for analysis. In such cases, the use of satellite data has various advantageous as these products have very large spatial coverage and are homogeneous across the globe, which make it easy to compare cloud fractions at various existing potential astronomical sites. Satellite data provides images over a hundred spectral bands which allow to diagnosis of several cloud characteristics such as cloud top height, liquid water content, cloud mask, brightness temperature, etc. (Wałaszek et al., 2017). The only limitation to the satellite data is due to poor temporal resolution.

Ground-based observations made by human eye can have high temporal resolution over a wider area (or nearly half a hemisphere) of the sky. However, such data dependent on experience of the observer as well as quality of the site. In this aspect, Indian Institute of Astrophysics (IIA), Bangalore, has carried out astronomical site survey program in the high-altitude Changthang region of eastern Ladakh for setting up astronomical facilities in optical and NIR regions from the early 1990s onward (Singh et al., 1989, Team, 1996, Bhatt et al., 2000). After performing several years of astronomical site survey work, 2-m Himalayan Chandra Telescope (HCT), operating in optical and near-infrared region, was installed at Indian Astronomical Observatory (IAO) Hanle ($32°{46}^{\prime }{46}^{\prime \prime }$N;$78°{57}^{\prime }{52}^{\prime \prime }$E; 4500 m, amsl) by IIA in 2000. It may be mentioned that the day and night sky log (cloud coverage) is maintained at IAO, Hanle from the late 1990s till date by telescope observing assistants for every one hour time interval. The regular scientific observations with the HCT have been initiated from May 2003 onwards. Several other additional astronomical facilities were also installed thereafter at IAO, Hanle and the details of the instruments and their scientific goals are described by Prabhu (2014).

In addition, there are several future plans for the development of solar as well as night-time astronomy at Hanle and regions nearby due to its promising site characteristics. For example, the Indian astronomical community has proposed for construction of a National Large Optical Telescope (NLOT) of 8–10 m class. During the recent past, IIA had also proposed development of National Large Solar Telescope (NLST) at Merak, which is around 117 km from Hanle and around 95 km far from Leh town (on aerial distance), Capital of Ladakh Union Territory. The present work is also a part of the evaluation of cloud coverage over the NLST site at Merak and NLOT site at Hanle using multi-year, homogeneous and continuous data. Fig. 1 shows the locations of sites of IAO-Hanle and Merak, with respect to Leh in Ladakh region.

In the present work, we used satellite and reanalysis data over the present location as well as several astronomical observatories across the globe to examine and compare the cloud coverage with Hanle and Merak sites. Using a common source of data will provides a better comparison at several sites across the globe. Since IAO, Hanle has long archival history of visual cloud coverage data, the present study also examined and compared with satellite data during 1997–2018. To our knowledge, such a study of cloud fraction from a homogeneous data source (visual observation) would be the largest temporal data coverage available at an astronomical observatory so far.

We organize this paper as follows: in Section 2, we describe the data coverage and methodology to define clear and useful nights in astronomical observation. In Section 3, we describe the meteorological conditions of the IAO sites using surface meteorological parameters. In Section 4, we present the results and discussion which describe the details of the cloud data from visual observation and its validation with Modern-Era Retrospective analysis for Research and Applications, version 2.0 (MERRA-2) and Atmospheric Infrared Sounder (AIRS) data, and subsequently results of comparative studies between Hanle, Merak and the best astronomical observatory sites in the world. Further, classification of clouds and precipitation data over the different observatories are also described in the sub-Section 4. Finally, Section 5 gives the conclusions of the present work.