CALIPSO Products Research Articles

Detecting Multilayer Clouds From the Geostationary Advanced Himawari Imager Using Machine Learning Techniques

This study develops a machine learning (ML)-based multilayer cloud detection algorithm for the passive Advanced Himawari Imager (AHI) aboard the geostationary Himawari-8 satellite. AHI measurements in the 0.64-, 1.6-, 2.3-, 3.9-, 7.3-, 8.6-, 11.2-, and 12.4- <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{m}$ </tex-math></inline-formula> channels, their combinations, geolocations, and observational geometries are used as predictors, and collocated active CloudSat and CALIPSO data are used to accurately label multilayer cloud pixels as the reference/truth of the predictand. We develop an ML-based daytime model (ML-Day) that utilizes all the aforementioned predictors and an all-time one (ML-All) that excludes the solar channel-dependent variables. Among four ML algorithms, the random forest (RF) performs slightly better than the artificial neural network, K-nearest neighbor, and support vector machines. By comparing with the merged CloudSat and CALIPSO product, the ML-Day model correctly identifies ~89% single-layer clouds and ~70% multilayer clouds, outperforming the Moderate Resolution Imaging Spectroradiometer (MODIS) operational multilayer cloud product (~80% and ~40% given by Marchant <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al.</i> ). The success rates of ML-All for single-layer and multilayer clouds also reach ~85% and ~64%, respectively. The misclassification of our algorithm is mostly caused by missing optically thin clouds, a drawback of most radiometers without the 1.38- <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$\mu \text{m}$ </tex-math></inline-formula> channel. Furthermore, with multilayer cloud pixels well detected by our algorithm, the AHI operational cloud top height retrievals are found to be larger biased due to multilayer cloud occurrence and might be improved by considering cloud vertical structures.

Effect of Dust Load on the Cloud Top Ice‐Water Partitioning Over Northern Middle to High Latitudes With CALIPSO Products

AbstractWe quantified effects of dust load on the cloud top ice cloud fraction (ICF) in terms of the dust extinction coefficient (σext). We analyzed 3‐year data sets obtained from an active satellite sensor over middle to high latitudes in the northern hemisphere for temperatures (T) between 230 and 273 K and σext values between 0.005 and 0.145 km−1. At about 250 K, ICF changed by about 30% in response to the above range of σext, whereas at extreme T values, ICF was relatively insensitive to σext. Thus, we concluded that ICF was primarily determined by T, with substantial influence of σext at about 250 K, likely due to increased opportunities for freezing as σext increases. Sensitivity of ICF was the lowest both at the largest σext and lowest T and at the smallest σext and highest T, while it was the highest at about 0.03 km−1 of σext and about 250 K.

Radiative Invigoration of Tropical Convection by Preceding Cirrus Clouds

AbstractThis work seeks evidence for convective–radiative interactions in satellite measurements, with a focus on the variability over the life cycle of tropical convection in search of the underlying processes at a fundamental level of the convective dynamics. To this end, the vertical profiles of cloud cover and radiative heating from the CloudSat–CALIPSO products are sorted into a composite time series around the hour of convective occurrence identified by the TRMM PR. The findings are summarized as follows. Cirrus cloud cover begins to increase, accompanied by a notable reduction of longwave cooling, in moist atmospheres even 1–2 days before deep convection is invigorated. In contrast, longwave cooling stays efficient and clouds remain shallow where the ambient air is very dry. To separate the radiative effects by the preceding cirrus clouds on convection from the direct effects of moisture, the observations with enhanced cirrus cover are isolated from those with suppressed cirrus under a moisture environment being nearly equal. It is found that rain rate is distinctly higher if the upper troposphere is cloudier regardless of moisture, suggesting that the cirrus radiative effects may be linked with the subsequent growth of convection. A possible mechanism to support this observational implication is discussed using a simple conceptual model. The model suggests that the preceding cirrus clouds could radiatively promote the moistening with the aid of the congestus-mode dynamics within a short period of time (about 2 days) as observed.

Seasonal Aerosol Vertical Distribution and Optical Properties over North China

To provide theoretical basis for studying the formation mechanism of aerosol, this paper analyzed the spatial and temporal variation characteristics of the aerosol optical depth (AOD), and the seasonal aerosol compositions, optical properties and vertical distribution under different polluted conditions over North China by using CALIPSO products. Then seasonal aerosol optical properties were validated by the AERONET data. The results show that the AOD at the plain regions is high while the AOD at mountainous regions is low, AOD in summer is the highest while AOD in spring is the lowest, and that AOD is minimum when the visibility is maximum. The AOD temporal variation during the recent ten years shows a weak increasing trend (0.003) over North China. The aerosols generated by human activities dominate over North China and the dust also plays an important role in spring, autumn and winter. The depolarization and color ratios are highest in spring and lowest in summer. The dust is transported to the middle and upper troposphere in spring, the strong vertical convection activities in summer transport massive aerosols from the atmospheric boundary layer to the free troposphere, and the stable meteorological conditions in autumn and winter trap massive aerosols within the boundary layer.

Using the NASA EOS A-Train to Probe the Performance of the NOAA PATMOS-x Cloud Fraction CDR

An important component of the AVHRR PATMOS-x climate date record (CDR)—or any satellite cloud climatology—is the performance of its cloud detection scheme and the subsequent quality of its cloud fraction CDR. PATMOS-x employs the NOAA Enterprise Cloud Mask for this, which is based on a naïve Bayesian approach. The goal of this paper is to generate analysis of the PATMOS-x cloud fraction CDR to facilitate its use in climate studies. Performance of PATMOS-x cloud detection is compared to that of the well-established MYD35 and CALIPSO products from the EOS A-Train. Results show the AVHRR PATMOS-x CDR compares well against CALIPSO with most regions showing proportional correct values of 0.90 without any spatial filtering and 0.95 when a spatial filter is applied. Values are similar for the NASA MODIS MYD35 mask. A direct comparison of PATMOS-x and MYD35 from 2003 to 2014 also shows agreement over most regions in terms of mean cloud amount, inter-annual variability, and linear trends. Regional and seasonal differences are discussed. The analysis demonstrates that PATMOS-x cloud amount uncertainty could effectively screen regions where PATMOS-x differs from MYD35.

Cumulus over the Tibetan Plateau in the Summer Based on CloudSat–CALIPSO Data

Abstract Cumulus (Cu) can transport heat and water vapor from the boundary layer to the free atmosphere, leading to the redistribution of heat and moist energy in the lower atmosphere. This paper uses the fine-resolution CloudSat–CALIPSO product to characterize Cu over the Tibetan Plateau (TP). It is found that Cu is one of the dominant cloud types over the TP in the northern summer. The Cu event frequency, defined as Cu occurring within 50-km segments, is 54% over the TP in the summer, which is much larger over the TP than in its surrounding regions. The surface wind vector converging at the central TP and the topographic forcing provide the necessary moisture and dynamical lifting of convection over the TP. The structure of the atmospheric moist static energy shows that the thermodynamical environment over the northern TP can be characterized as having weak instability, a shallow layer of instability, and lower altitudes for the level of free convection. The diurnal variation of Cu with frequency peaks during the daytime confirms the surface thermodynamic control on Cu formation over the TP. This study offers insights into how surface heat is transported to the free troposphere over the TP and provides an observational test of climate models in simulating shallow convection over the TP.

3D Structure of Saharan Dust Transport Towards Europe as Seen by CALIPSO

We present a 3D multi-year monthly mean climatology of Saharan dust advection over Europe using an area-optimized pure dust CALIPSO product. The product has been developed by applying EARLINET-measured dust lidar ratios and depolarization-based dust discrimination methods and it is shown to have a very good agreement in terms of AOD when compared to AERONET over Europe/North Africa and MODIS over Mediterranean. The processing of such purely observational data reveals the certain seasonal patterns of dust transportation towards Europe and the Atlantic Ocean. The physical and optical properties of the dust layer are identified for several areas near the Saharan sources, over the Mediterranean and over continental Europe.

Next-generation angular distribution models for top-of-atmosphere radiative flux calculation from CERES instruments: validation

Abstract. Radiative fluxes at the top of the atmosphere (TOA) from the Clouds and the Earth's Radiant Energy System (CERES) instrument are fundamental variables for understanding the Earth's energy balance and how it changes with time. TOA radiative fluxes are derived from the CERES radiance measurements using empirical angular distribution models (ADMs). This paper evaluates the accuracy of CERES TOA fluxes using direct integration and flux consistency tests. Direct integration tests show that the overall bias in regional monthly mean TOA shortwave (SW) flux is less than 0.2 Wm−2 and the RMSE is less than 1.1 Wm−2. The bias and RMSE are very similar between Terra and Aqua. The bias in regional monthly mean TOA LW fluxes is less than 0.5 Wm−2 and the RMSE is less than 0.8 Wm−2 for both Terra and Aqua. The accuracy of the TOA instantaneous flux is assessed by performing tests using fluxes inverted from nadir- and oblique-viewing angles using CERES along-track observations and temporally and spatially matched MODIS observations, and using fluxes inverted from multi-angle MISR observations. The averaged TOA instantaneous SW flux uncertainties from these two tests are about 2.3 % (1.9 Wm−2) over clear ocean, 1.6 % (4.5 Wm−2) over clear land, and 2.0 % (6.0 Wm−2) over clear snow/ice; and are about 3.3 % (9.0 Wm−2), 2.7 % (8.4 Wm−2), and 3.7 % (9.9 Wm−2) over ocean, land, and snow/ice under all-sky conditions. The TOA SW flux uncertainties are generally larger for thin broken clouds than for moderate and thick overcast clouds. The TOA instantaneous daytime LW flux uncertainties derived from the CERES-MODIS test are 0.5 % (1.5 Wm−2), 0.8 % (2.4 Wm−2), and 0.7 % (1.3 Wm−2) over clear ocean, land, and snow/ice; and are about 1.5 % (3.5 Wm−2), 1.0 % (2.9 Wm−2), and 1.1 % (2.1 Wm−2) over ocean, land, and snow/ice under all-sky conditions. The TOA instantaneous nighttime LW flux uncertainties are about 0.5–1 % (&lt; 2.0 Wm−2) for all surface types. Flux uncertainties caused by errors in scene identification are also assessed by using the collocated CALIPSO, CloudSat, CERES and MODIS data product. Errors in scene identification tend to underestimate TOA SW flux by about 0.6 Wm−2 and overestimate TOA daytime (nighttime) LW flux by 0.4 (0.2) Wm−2 when all CERES viewing angles are considered.

Characterizing a persistent Asian dust transport event: Optical properties and impact on air quality through the ground-based and satellite measurements over Nanjing, China

The optical properties, time-height distribution and impact on the local air quality from a heavy Asian dust transport episode are investigated with a synergistic ground-based, satellite sensors and transport model on 1 May, 2011 at Nanjing (32.05° N, 118.78° E, and 94 m ASL) in southeast China. Two dust layers located in the planetary-boundary-layer (PBL, <2.5 km) and free troposphere (3–6 km) are observed by a Polarization Raman-Mie Lidar, with the lower one originating from the Gobi deserts and the higher one from the Taklimakan deserts. The dust aerosol layer shows the depolarization ratios at 0.1–0.2 and strong extinction coefficients of 1.0 km−1 at 532-nm, while the extinction-to-backscatter ratios (e.g. lidar ratios) of dust are 47.3–55 sr below 2.5 km. During this dust intrusion period, the aerosol optical depths (AOD) dramatically increase from 0.7 to 1.6 at 500-nm whereas the Angstrom exponents decrease from 1.2 to 0.2. Meanwhile, surface PM10 and PM2.5 concentrations show a significant and coincident increase with the peak value reaching 767 μg/m3 and 222 μg/m3, respectively, indicating the mixture of dust with the anthropogenic aerosols. Regional influences of the transported dust in east China are further illustrated by the AERONET-sunphotometer at Taihu and Xianghe sites (downwind and upwind from Nanjing), satellites MODIS, CALIPSO and model products. Furthermore, the model product of dust profile and surface concentration are evaluated with the ground-based and CALISPO observation. The results indicate the model is capable of simulating the right timing of dust transport event and most loading below 3-km altitude; normalization of model dust with the PM10 near the Gobi deserts improves modeling surface dust concentration in Nanjing.

Airborne verification of CALIPSO products over the Amazon: a case study of daytime observations in a complex atmospheric scene

Abstract. A daytime underflight of CALIPSO with the Facility for Airborne Atmospheric Measurements was performed on 20 September 2012 in the Amazon region of Brazil, during the biomass burning season. The scene is dominated by a thin elevated layer (aerosol optical depth (AOD) 0.03 at 532 nm) and a moderately turbid boundary layer (AOD ~ 0.2 at 532 nm). The boundary layer is topped with small broken stratocumulus clouds. In this complex scene, a comparison of observations from the airborne and spaceborne lidars reveals a few discrepancies. The CALIPSO detection scheme tends to miss the elevated thin layer, and also shows several gaps (~ 30%) in the boundary layer. The small clouds are not correctly removed from the signals; this can cause the CALIPSO aerosol subtype to oscillate between smoke and polluted dust and may introduce distortion in the aerosol retrieval scheme. The magnitude of the average extinction coefficient estimated from CALIPSO Level 2 data in the boundary layer is as expected, when compared to the aircraft lidar and accounting for wavelength scaling. However, when the gaps in aerosol detection mentioned above are accounted for, we are left with an overall estimate of AOD for this particular scene that is of the order of two thirds of that determined with the airborne lidar.

Assessment of CALIPSO attenuated backscatter and aerosol retrievals with a combined ground-based multi-wavelength lidar and sunphotometer measurement

CALIPSO Level-1 attenuated backscatter and Level-2 aerosol products (Version-3.01) are evaluated with a combined ground-based lidar and AERONET-sunphotometer measurements in the daytime over the New York metropolitan area. To assess the CALIPSO Level-1 product, we combine the co-located ground-lidar and sunphotometer to derive aerosol extinction and backscatter profiles, and then simulate the CALIPSO equivalent attenuated backscatter coefficients. Direct statistical comparisons show a strong correlation (R = 0.92) and modest relative errors. Both dust and smoke plume events are focused to evaluate CALIPSO Level-2 aerosol layer products. The CALIPSO algorithms for cloud-aerosol discrimination and aerosol type classification are shown for the most part to work well, with a few exceptions in cases of aloft plumes with high aerosol loading. Small partitions of dense smoke are misclassified as the clouds or polluted dusts in the CALIPSO product. The aerosol extinction and backscatter coefficients are generally consistent between the CALIPSO and ground-based retrievals, but both geometric thickness and column optical depths of aerosol layers from CALIPSO products are underestimated. In addition, we find that some weakly scattering aerosol layers are clearly displayed by the ground-based lidar, but not identified by the current CALIPSO algorithm due to the detection sensitivity issue.

The depolarization–attenuated backscatter relationship for dust plumes

This study identified the relationship between the layer-integrated attenuated backscatter coefficient and layer-integrated depolarization ratio of dust plumes and compared it with that of cloud, using CALIPSO LIDAR measurements. The histogram distribution of the integrated color ratio for dust and cloud was also examined. On the basis of the layer-integrated attenuated backscatter coefficient and layer-integrated depolarization ratio relation, a simple method of detecting dust plumes was developed. A case study of dust identification over the Taklimakan Desert was conducted and compared with the current CALIPSO products. The result shows that the proposed method can significantly improve the classification of cloud and dust plumes and can supplement the current space-borne LIDAR discrimination approach, especially over dust source regions. In addition, The zonal and meridional mean occurrence derived by the proposed method and the CALIPSO's method were compared for Asian dust over East Asia region (30°N -45°N, 80°E -180°E) using the night measurements of CALIPSO from March to May, 2007. The comparison showed that the dust occurrence obtained from the proposed method is larger than that of CALIPSO's method. The dust could be found up to around 6-8 km (Above Sea Level, ASL) near the Taklimakan desert region, and maximum occurrence is over 80%. The transport altitude remained at 3 km-7 km (ASL) as the dust was transported across the Pacific Ocean.

Aerosol indirect effect on warm clouds over South-East Atlantic, from co-located MODIS and CALIPSO observations

Abstract. In this study, we provide a comprehensive analysis of aerosol interaction with warm boundary layer clouds over the South-East Atlantic. We use aerosol and cloud parameters derived from MODIS observations, together with co-located CALIPSO estimates of the layer altitudes, to derive statistical relationships between aerosol concentration and cloud properties. The CALIPSO products are used to differentiate between cases of mixed cloud-aerosol layers from cases where the aerosol is located well-above the cloud top. This technique allows us to obtain more reliable estimates of the aerosol indirect effect than from simple relationships based on vertically integrated measurements of aerosol and cloud properties. Indeed, it permits us to somewhat distinguish the effects of aerosol and meteorology on the clouds, although it is not possible to fully ascertain the relative contribution of each on the derived statistics. Consistently with the results from previous studies, our statistics clearly show that aerosol affects cloud microphysics, decreasing the Cloud Droplet Radius (CDR). The same data indicate a concomitant strong decrease in cloud Liquid Water Path (LWP), which is inconsistent with the hypothesis of aerosol inhibition of precipitation (Albrecht, 1989). We hypothesise that the observed reduction in LWP is the consequence of dry air entrainment at cloud top. The combined effect of CDR decrease and LWP decrease leads to rather small sensitivity of the Cloud Optical Thickness (COT) to an increase in aerosol concentration. The analysis of MODIS-CALIPSO coincidences also evidences an aerosol enhancement of low cloud cover. Surprisingly, the Cloud Fraction (CLF) response to aerosol invigoration is much stronger when (absorbing) particles are located above cloud top than in cases of physical interaction. This result suggests a relevant aerosol radiative effect on low cloud occurrence: absorbing particles above the cloud top may heat the corresponding atmosphere layer, decrease the vertical temperature gradient, increase the low tropospheric stability and provide favourable conditions for low cloud formation. We also analyse the impact of anthropogenic aerosols on precipitation, through the statistical analysis of CDR-COT co-variations. A COT value of 10 is found to be the threshold beyond which precipitation is mostly formed, in both clean and polluted environments. For larger COT, polluted clouds show evidence of precipitation suppression. Results suggest the presence of two competing mechanisms governing LWP response to aerosol invigoration: a drying effect due to aerosol enhanced entrainment of dry air at cloud top (predominant for optically thin clouds) and a moistening effect due to aerosol inhibition of precipitation (predominant for optically thick clouds).

Size matters: Influence of multiple scattering on CALIPSO light‐extinction profiling in desert dust

We investigate the discrepancies in measurements of light extinction and extinction‐to‐backsatter ratio (lidar ratio) of desert dust with CALIPSO and ground‐based lidar systems. Multiwavelength polarization Raman lidar measurements in the Saharan dust plume performed at Praia, Cape Verde, 15.0°N, 23.5°W, during SAMUM–2 in June 2008 were analyzed and compared to results of nearby CALIPSO overflights. The particle extinction coefficients and thus the optical depth are underestimated in the CALIPSO products by about 30% compared to Raman lidar measurements. A pre‐defined lidar ratio of 40 sr at 532 nm is used for mineral dust in the CALIPSO algorithms in agreement with values of 41 ± 6 sr found from constrained retrievals. However, the ground‐based lidar observations show much larger values of the order of 55 ± 10 sr. The discrepancies can be explained by the influence of multiple scattering which is ignored in the CALIPSO retrievals. Based on recent observations of the size distribution of dust particles from airborne in‐situ observations during SAMUM–1, our model calculations show that the multiple‐scattering‐related underestimation of the extinction coefficient in the CALIPSO lidar signals ranges from 10%–40%. We propose a method to overcome this underestimation.