Investigations into the development of a satellite-based aerosol climate data record using ATSR-2, AATSR and AVHRR data over north-eastern China from 1987 to 2012

Abstract

Abstract. Satellites provide information on the temporal and spatial distributions of aerosols on regional and global scales. With the same method applied to a single sensor all over the world, a consistent data set is to be expected. However, the application of different retrieval algorithms to the same sensor and the use of a series of different sensors may lead to substantial differences, and no single sensor or algorithm is better than any other everywhere and at all times. For the production of long-term climate data records, the use of multiple sensors cannot be avoided. The Along Track Scanning Radiometer (ATSR-2) and the Advanced ATSR (AATSR) aerosol optical depth (AOD) data sets have been used to provide a global AOD data record over land and ocean of 17 years (1995–2012), which is planned to be extended with AOD retrieved from a similar sensor. To investigate the possibility of extending the ATSR data record to earlier years, the use of an AOD data set from the Advanced Very High Resolution Radiometer (AVHRR) is investigated. AOD data sets used in this study were retrieved from the ATSR sensors using the ATSR Dual View algorithm ADV version 2.31, developed by Finnish Meteorological Institute (FMI), and from the AVHRR sensors using the aerosol optical depth over land (ADL) algorithm developed by RADI/CAS. Together, these data sets cover a multi-decadal period (1987–2012). The study area includes two contrasting areas, both in regards to aerosol content and composition and surface properties, i.e. a region over north-eastern China, encompassing a highly populated urban/industrialized area (Beijing–Tianjin–Hebei) and a sparsely populated mountainous area. Ground-based AOD observations available from ground-based sun photometer AOD data in AERONET and CARSNET are used as a reference, together with broadband extinction method (BEM) data at Beijing to cover the time before sun photometer observations became available in the early 2000s. In addition, MODIS-Terra C6.1 AOD data are used as a reference data set over the wide area where no ground-based data are available. All satellite data over the study area were validated against the reference data, showing the qualification of MODIS for comparison with ATSR and AVHRR. The comparison with MODIS shows that AVHRR performs better than ATSR in the north of the study area (40∘ N), whereas further south ATSR provides better results. The validation against sun photometer AOD shows that both AVHRR and ATSR underestimate the AOD, with ATSR failing to provide reliable results in the wintertime. This is likely due to the highly reflecting surface in the dry season, when AVHRR-retrieved AOD traces both MODIS and reference AOD data well. However, AVHRR does not provide AOD larger than about 0.6 and hence is not reliable when high AOD values have been observed over the last decade. In these cases, ATSR performs much better for AOD up to about 1.3. AVHRR-retrieved AOD compares favourably with BEM AOD, except for AOD higher than about 0.6. These comparisons lead to the conclusion that AVHRR and ATSR AOD data records each have their strengths and weaknesses that need to be accounted for when combining them in a single multi-decadal climate data record.

Air Quality over China

“…To this end, the RADI/CAS aerosol optical depth over land (ADL) algorithm for AVHRR [72] was further developed. The evaluation of the AVHRR-derived AOD versus AERONET data and versus AOD derived from solar radiation measurements, using the broadband extinction method (BEM), shows the good performance for low AOD, but improvements are needed for high AOD [40].…”

“…To this end, the RADI/CAS aerosol optical depth over land (ADL) algorithm [72] was further developed and evaluated versus AERONET and CARSNET data available over the study area. However, these reference data were not available before 2000 and therefore AOD derived from solar radiation measurements, using the broadband extinction method (BEM) as briefly described in the Introduction section, were used for validation [40].…”

“…A primary focus of the AQ-China aerosol project was the use of Aerosol_cci AOD data to present the current status of aerosol concentrations over China. To this end, ATSR data were used and, in view of the time coverage from 1995 until 2012, the data were extended both backward to 1987 using AVHRR data [40] and forward to 2017 [4,23]. The results were described in detail in the indicated references and will not be repeated here.…”

“…The work by de Leeuw et al [4] and Sogacheva et al [23,39] on the spatiotemporal variation of the aerosol optical depth (AOD) over China for the period 1995-2017, including the use of ATSR and MODIS data to construct an AOD time series, is extended to 2020. As part of the Dragon 4 cooperation, the ATSR time series was extended back in time to 1983 using AVHRR data [40]. MODIS data was also used together with CALIOP data to study the 3-D variation of aerosol properties over the Beijing-Tianjin-Hebei area (BTH), the Yangtze River Delta (YRD) and the Pearl River Delta (PRD) [41].…”

Meteorological and Ancillary Data Resources for Climate Research in Urban Areas

“…NOAA's Advanced Very High-Resolution Radiometer (AVHRR) two and three instruments have retrieved LST at 1.1 km resolution since 1981. Early studies of the SUHI were based on AVHRR data [76][77][78], and the datasets are still used [79].…”