Uncertainty of AMSU-A derived temperature trends in relationship with clouds and precipitation over ocean
- First Online:
- Cite this article as:
- Weng, F., Zou, X. & Qin, Z. Clim Dyn (2014) 43: 1439. doi:10.1007/s00382-013-1958-7
- 268 Views
Microwave Sounding Unit (MSU) and Advanced Microwave Sounding Unit-A (AMSU-A) observations from a series of National Oceanic and Atmospheric Administration satellites have been extensively utilized for estimating the atmospheric temperature trend. For a given atmospheric temperature condition, the emission and scattering of clouds and precipitation modulate MSU and AMSU-A brightness temperatures. In this study, the effects of the radiation from clouds and precipitation on AMSU-A derived atmospheric temperature trend are assessed using the information from AMSU-A window channels. It is shown that the global mean temperature in the low and middle troposphere has a larger warming rate (about 20–30 % higher) when the cloud-affected radiances are removed from AMSU-A data. It is also shown that the inclusion of cloud-affected radiances in the trend analysis can significantly offset the stratospheric cooling represented by AMSU-A channel 9 over the middle and high latitudes of Northern Hemisphere.
The Microwave Sounding Unit (MSU) and the Advanced Microwave Sounding Unit-A (AMSU-A) on board the National Oceanic and Atmospheric Administration (NOAA) polar-orbiting satellites measure the atmospheric radiation from 23 to 89 GHz. Under a clear-sky atmospheric condition, the radiative energy primarily comes from the oxygen emission. Since the oxygen concentration is nearly uniformly distributed through the Earth’s atmosphere and is stable with time, MSU and AMSU-A are unique satellite instruments for remotely sounding the atmospheric temperature. The MSU instruments on board Tiros-N, NOAA-6 to NOAA-14 have four channels and provided data from 1979 to 2006. Each of the four channels provides measurements of a weighted average of radiation emitted from a particular layer of the atmosphere at a specified frequency. The relative contributions to the total measured radiance from different levels of the atmosphere are quantified by the so-called weighting function (WF), which is channel-dependent. The measured radiation is most sensitive to the atmospheric temperature at the altitude where WF reaches the maximum value. The AMSU-A instruments on board NOAA-15 to NOAA-19 have 15 channels, among which four channels (i.e., AMSU-A channels 3, 5, 7, and 9) are similar, but not identical, to the four MSU channels in frequency. The other 11 AMSU-A channels sample more atmospheric layers than MSU. By putting together MSU and the MSU-like AMSU-A channels, a long-term series of global satellite microwave sounding data of more than 30 years is available for climate study related to atmospheric temperature changes (Zou et al. 2009; Wang et al. 2012).
The atmospheric temperature trends derived from the MSU and AMSU-A on board the NOAA polar-orbiting satellites have been a subject of debate. Pioneer investigations by Spencer and Christy (1992a, b) and their follow-on work at the University of Alabama at Huntsville (UAH) (Christy et al. 1998, 2000, 2003) showed nearly no warming trends for the mid-tropospheric temperature time series derived from the MSU channel 2 (53.74 GHz) and AMSU-A channel 5 (53.71 GHz) observations (called T2). However, the Remote Sensing Systems (RSS) (Mears et al. 2003; Mears and Wentz 2009) and University of Maryland (Vinnikov et al. 2006) and NOAA/NESDIS/Center for Satellite Applications and Research (STAR) (Zou et al. 2009) groups obtained a small warming trend from the same satellite observations. The most recent analysis of different datasets shows a global ocean mean T2 trend of 0.080 ± 0.103 K decade−1 for UAH, 0.135 ± 0.113 K decade−1 for RSS, 0.22 ± 0.07 K decade−1 for UMD and 0.200 ± 0.067 K decade−1 for STAR for the time period from 1987 to 2006 (Zou et al. 2009). In the past, the analysis of the MSU and AMSU-A decadal trend focused on the following areas: (1) diurnal adjustment of various instruments; (2) correction of earth incidence angle; and (3) inter-calibration of co-orbiting satellite (Mears and Wentz 2009). Qin et al. (2012) pointed out a nonlinear behavior of the decadal climate trends of most AMSU-A channels using data from NOAA-15 satellite over the time period from October 26, 1998 to August 7, 2010.
A concern earlier was also raised on how much of the MSU signal arises from non-oxygen emission. For example, for MSU channel 2, Spencer et al. (1990) predicted a small influence of contamination from precipitation-sized ice in deep convection, cloud water, water vapor and surface emissivity. It was concluded that the largest effects come from precipitation-sized ice in deep convection, causing a brightness temperature depression of up to several degrees Celsius. Therefore, the MSU data have been filtered to remove this particular contamination. Spencer (1993) also developed a technique for calculating the anomalous temperature increase in MSU channel 1 (50.3 GHz) and then calibrating it to a rainfall rate. Prabhakara et al. (1995, 1996) suggested that a substantial hydrometeor effect exists in the MSU records, which was criticized by Spencer et al. (1996) who argued that the residual hydrometeor contamination effects were greatly overestimated by Prabhakara et al. (1995, 1996).
This study further evaluates the impacts of both cloud and precipitation on AMSU-A sounding channels and investigates the uncertainty of tropospheric and low stratospheric temperature trends with and without their contributions. A radiative transfer model is used for illustrating the responses of AMSU-A brightness temperatures to several cloud parameters in Sect. 2. Numerical results of cloud impact on trend are provided in Sect. 3 using NOAA-15 AMSU-A observations. Summary and conclusions are provided in Sect. 4.
2 AMSU-A brightness temperature data
The brightness temperature observations used in this study are from the first AMSU-A instrument on board the NOAA Advanced TOVS (ATOVS) system, NOAA-15, which was launched on May 13, 1998. The NOAA-15 satellite has an inclination angle of 98.7°, a local equator crossing time (LECT) of about 7:30 pm in its ascending node at its launch time, and an altitude of 833 km above the Earth.
Channel characteristics of AMSU-A channels 3, 5, 7 and 9
Band width (MHz)
Beam width (°)
Weighting function peak (hPa)
53.596 ± 0.115
The emission and scattering of clouds and precipitation modulate brightness temperatures from both MSU and AMSU-A. In order to assess the effects of the radiation from clouds and precipitation on atmospheric temperature trend derived from microwave temperature sounding instruments (e.g., MSU and AMSU-A), cloud-affected brightness temperature measurements must be identified. Fortunately, the information from AMSU-A window channels 1 and 2 can be used for retrieving atmospheric cloud liquid water path (LWP) using the algorithm developed by Weng et al. (2003). For this reason, temperature trends for AMSU-A channels 3, 5, 7 and 9 from NOAA-15 are investigated in this study. Uncertainty of MSU derived temperature trends in relationship with clouds and precipitation requires a new algorithm to be developed for detecting cloud-affected MSU brightness temperature observations, which is planned for future research. Until then, cloud impacts on atmospheric temperature trends derived from a much longer climate data record of microwave temperature sounding data could be assessed.
3 An idealized sensitivity study
Here, the Community Radiative Transfer Model (CRTM) is first used for simulating the responses of AMSU-A brightness temperatures to the clouds characterized by rain rate. The CRTM was developed by the US Joint Center for Satellite Data Assimilation (JCSDA) for rapid calculations of satellite radiances and their derivatives under various atmospheric and surface conditions (Weng 2007; Han et al. 2007). It supports a large number of sensors, including the historical and near future sensors from the Joint Polar Satellite System (JPSS) and the Geostationary Operational Environmental Satellite-R Series (GOES-R), covering the microwave, infrared and visible wavelengths for a variety of research and applications. For calculations of cloud and aerosol absorption and scattering, lookup tables of the optical properties of six cloud and eight aerosol types are included in the cloud/aerosol optical property module. The fast doubling-adding method is used to solve the multi-stream radiative transfer equation.
An idealized cloud distribution similar to that in Liu and Curry (1993) is assumed for the sensitivity study. Namely, a 0.8-km non-precipitating cloud layer with a liquid water path (LWP) of 0.5 kg m−2 is added to the atmosphere below the freezing level and a precipitation layer is set below the cloud. Each specific rain rate is assumed for the entire precipitation layer. A schematic diagram of the cloud and precipitation system is illustrated in Fig. 1. Note that an atmospheric profile including temperature and water vapor is used as inputs to CRTM and the surface emissivity is set to 0.5, which is a value close to an ocean surface.
Based on the results in Figs. 1, 2, it is concluded that the larger the raindrop size is, the larger the scattering and emission effects of clouds on the brightness temperatures are. The cloud emission dominates when the rain rate is low and the raindrop size is small. In reality, clouds are of different types and are located at different altitudes. Only the integrated effect of clouds on the temperature trend can be estimated from real data.
4 Trend analysis
The global impact of clouds on temperature trends in the upper troposphere and low stratosphere are negligible. If we separate Northern Hemisphere from Southern Hemisphere, impacts of clouds on temperature trends become more significant than the global average. In the Southern Hemisphere, the warming trends for both the window channel 3 and the middle-tropospheric channel 5 are 0.06 K decade−1 and 0.03 K decade−1, respectively, which decrease to nearly zero when clouds are present (Fig. 5b). The temperature cooling trends for channels 7 and 8 are slightly increased by clouds (Fig. 5b) in the Southern Hemisphere. Cloud impact on temperature trend for channel 5 in the Northern Hemisphere is similar to that in the Southern Hemisphere, i.e., cloud reduces the warming trend. Impact of clouds on channel 7 is also negligible in the Northern Hemisphere. However, an opposite effect of cloud on temperature trend is found in the Northern Hemisphere near the surface and in the upper troposphere and low stratosphere: clouds causes an overestimate of temperature warm trend of the middle-tropospheric channel 5 and an underestimate of temperature cooling trend of the upper troposphere and low stratosphere channel 9. The global ocean mean trends for channel 5 within the time period from October 26, 1998 to August 7, 2010 with or without removing cloud-affected data obtained in this study are much smaller than those obtained by UAH, UMD and STAR for the time period from 1987 to 2006 (Zou et al. 2009). Further investigation is needed to clarify if the different trends can be caused by the data record length.
5 Summary and conclusions
In this study, the cloudy and cloud-free atmospheric temperature trends are derived from satellite microwave temperature sounding observations. The global daily mean brightness temperatures observed by the NOAA polar-orbiting satellite NOAA-15 AMSU-A instrument over the 13-year time period from August 1, 1999 to June 30, 2012 over ocean are used in our analysis. The traditional linear regression method is applied to the 13-years global brightness temperatures of AMSU-A channels 3, 5, 7 and 9 to obtain global and hemispheric averaged warming and cooling trends. By binning all AMSU-A channels brightness temperature measurements within five-degree latitudinal bands, the latitudinal dependence of the cloud impact on global warming trend is calculated and its relationship to global average warming/cooling is determined. It is shown that the atmospheric warming trends in middle latitudes are significantly larger when cloud effects are removed from the 13-year microwave sounding data of AMSU-A channels 3 and 5 in both hemispheres. The scattering and emission effect of clouds and precipitation significantly reduces the values of the warm trends in the low and middle troposphere derived from the microwave data. The cooling trends are found in the high latitudes in the Southern Hemisphere and all latitudes in Northern Hemisphere. The high altitude clouds tend to reduce the cooling trends of AMSU-A channel 9 in Northern Hemisphere, especially in middle and high latitudes in Northern Hemisphere. The cloud impacts on the cooling trends in Southern Hemisphere are negligible. However, their impacts on the temperature trends could be much larger in latitudinal zones than that on the global warm trend.
In the present study, only a single satellite data (NOAA-15 AMSU-A) is used for the investigation of the cloud impact on temperature trends. It is recommended that a similar study be carried out on a longer time series (e.g., more than 30 years) by combining MSU data with AMSU-A data from multi-satellite data. However, the cloud algorithm for AMSU-A cannot be applied to MSU data since the two AMSU-A window channels used for the LWP retrieval are not available from MSU data. Further investigation is required for developing an approximate cloud algorithm for MSU data before the cloud impact on temperature trends can be assessed from MSU data.
This work was supported by Chinese Ministry of Science and Technology under 973 project 2010CB951600.