RNM119 - Artículos

Aerosol type classification with machine learning techniques applied to multiwavelength lidar data from EARLINET

Aerosol type classification with machine learning techniques applied to multiwavelength lidar data from EARLINET del Águila, Ana; Ortiz Amezcua, Pablo; Tabik, Siham; Bravo Aranda, Juan Antonio; Fernández Carvelo, Sol; Alados Arboledas, Lucas

Study of the exceptional meteorological conditions, trace gases and particulate matter1 measured during the 2017 forest fire in Doñana Natural Park, Spain

Study of the exceptional meteorological conditions, trace gases and particulate matter1 measured during the 2017 forest fire in Doñana Natural Park, Spain Adame, J.A.; Lope, L.; Hidalgo, P.J.; Sorribas, M.; Gutiérrez-Álvarez, I.; del Águila, Ana; Saiz-Lopez, A.; Yela, M. In late June 2017, a forest fire occurred in Doñana Natural Park, which is located in southwestern Europe. Many animal and plant species, some of which are threatened, suffered from the impact of this fire, and important ecosystems in the European Union were seriously affected. This forest fire occurred under exceptional weather conditions. The meteorological situation was studied at both the synoptic scale and the local scale using meteorological fields in the ERA-Interim global model from ECMWF (European Centre for Medium Range Weather Forecasts), the WRF (Weather Research and Forecasting) mesoscale model and ground observations collected at El Arenosillo observatory. Anomalies were obtained using records (observations and simulations) over the last two decades (1996–2016). An anticyclonic system dominated the synoptic meteorological conditions, but a strong pressure gradient was present; positive high pressure anomalies and negative low pressure anomalies resulted in intense NW flows. At the surface, wind gusts of 80 km h−1, temperatures up to 35 °C and relative humidity values <20% were observed. In terms of anomalies, these observations corresponded to positive temperature anomalies (differences of 12 °C), positive wind speed anomalies (>29 km h−1) and negative relative humidity anomalies (differences of 40%). The forest fire reached El Arenosillo observatory approximately 8 h after it began. When the fire started, record-setting maximum values were measured for all gases monitored at this site (specifically, peaks of 99,995 μg m−3 for CO, 951 μg m−3 for O3, 478 μg m−3 for NO2, 116 μg m−3 for SO2 and 1000 μg m−3 for PM10). According to the temporal evolution patterns of these species, the atmosphere over a burnt area can recover to initial atmospheric levels between 48 and 96 h after an event. The impact of the Doñana plume was studied using hourly forward trajectories computed with the HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) model to analyse the emission source for the burnt area. The Doñana fire plume affected large metropolitan areas near the Mediterranean coast. Air quality stations located in the cities of Seville and Cadiz registered the arrival of the plume based on increases in CO and PM10. Using CO as a tracer, measurements from the AIRS and MOPITT instruments allowed us to observe the transport of the Doñana plume from the Strait of Gibraltar to the Mediterranean. Finally, after two days, the Doñana forest fire plume reached the western Mediterranean basin.

Climatological study for understanding the aerosol radiative effects at southwest Atlantic coast of Europe

Climatological study for understanding the aerosol radiative effects at southwest Atlantic coast of Europe Sorribas, Mar; Andrews, Elisabeth; Ogren, J.A.; del Águila, Ana; Fraile, R.; Sheridan, P.; Yela, Margarita In order to describe the means, variability and trends of the aerosol radiative effects on the southwest Atlantic coast of Europe, 11 years of aerosol light scattering (σsp) and 4 years of aerosol light absorption (σap) are analyzed. A 2006–2016 trend analysis of σsp for D < 10 μm indicates statistically significant trends for March, May–June and September–November, with a decreasing trend ranging from −1.5 to −2.8 Mm−1/year. In the 2009–2016 period, the decreasing trend is only observed for the months of June and September. For scattering Ångström exponent (SAE) there is an increasing trend during June with a rate of 0.059/year and a decreasing trend during October with −0.060/year. The trends observed may be caused by a reduction of Saharan dust aerosol or a drop in particle loading in anthropogenic influenced air masses. The relationship between SAE and absorption Ångström exponent is used to assess the aerosol typing. Based on this typing, the sub-micron particles are dominated by black carbon, mixed black and brown carbon or marine with anthropogenic influences, while the super-micrometer particles are desert dust and sea spray aerosol. The mean and standard deviation of the dry aerosol direct radiative effect at the top of the atmosphere (DRETOA) are −4.7 ± 4.2 W m−2. DRETOA for marine aerosol shows all observations more negative than −4 W m−2 and for anthropogenic aerosol type, DRETOA ranges from −5.0 to −13.0 W m−2. DRETOA of regional marine aerosol ranges from −3 to −7 W m−2, as it consists of a mixture of sea salt and anthropogenic aerosol. The variability in DRETOA is mainly dependent on AOD, given that variations in backscatter fraction and the single scattering albedo tend to counteract each other in the radiative forcing efficiency equation. The results shown here may help in interpretation of satellite retrieval products and provide context for model evaluation.

Cirrus-induced shortwave radiative effects depending on their optical and physical properties: Case studies using simulations and measurements

Cirrus-induced shortwave radiative effects depending on their optical and physical properties: Case studies using simulations and measurements Córdoba-Jabonero, Carmen; Gómez-Martin, Laura; del Águila, Ana; Vilaplana, J.M.; López Cayuela, María Ángeles; Zorzano-Mier, María Paz Cirrus (Ci) clouds play an important role in the atmospheric radiative balance, and hence in Climate Change. In this work, a polarized Micro-Pulse Lidar (P-MPL), standard NASA/Micro Pulse NETwork (MPLNET) system, deployed at the INTA/El Arenosillo station in Huelva (SW Iberian Peninsula) is used for Ci detection and characterization for the first time at this site. Three days were selected on the basis of the predominantly detected Ci clouds in dependence on their cloud optical depth (COD). Hence, three Ci cloud categories were examined at day-times for comparison with solar radiation issues: 19 cases of sub-visuals (svCi, COD: 0.01–0.03) on 1 October 2016, 7 cases of semitransparents (stCi, COD: 0.03–0.30) on 8 May 2017, and 17 cases of opaques (opCi, COD: 0.3–3.0) on 28 October 2016. Their radiative-relevant optical, macro- and micro-physical properties were retrieved. The mean COD for the svCi, stCi and opCi groups was 0.02 ± 0.01, 0.22 ± 0.08 and 0.93 ± 0.40, respectively; in overall, their lidar ratio ranged between 25 and 35 sr. Ci clouds were detected at 11–13 km height (top boundaries) with geometrical thicknesses of 1.7–2.0 km. Temperatures reported at those altitudes corresponded to lower values than the thermal threshold for homogenous ice formation. Volume linear depolarization ratios of 0.3–0.4 (and normalized backscattering ratios higher than 0.9) also confirmed Ci clouds purely composed of ice particles. Their effective radius was within the interval of 9–15 μm size, and the ice water path ranged from 0.02 (svCi) to 9.9 (opCi) g m−2. The Cirrus Cloud Radiative Effect (CCRE) was estimated using a Radiative Transfer (RT) model for Ci-free conditions and Ci-mode (Ci presence) scenarios. RT simulations were performed for deriving the CCRE at the top-of-atmosphere (TOA) and on surface (SRF), and also the atmospheric CCRE, for the overall shortwave (SW) range and their spectral sub-intervals (UV, VIS and NIR). A good agreement was first obtained for the RT simulations as validated against solar radiation measurements under clean conditions for solar zenith angles less than 75° (differences were mainly within ±20 W m−2 and correlation coefficients close to 1). By considering all the Ci clouds, independently on their COD, the mean SW CCRE values at TOA and SRF were, respectively, −30 ± 26 and − 24 ± 19 W m−2, being the mean atmospheric CCRE of −7 ± 7 W m−2; these values are in good agreement with global annual estimates found for Ci clouds. By using linear regression analysis, a Ci-induced enhancing cooling radiative effect was observed as COD increased for all the spectral ranges, with high correlations. In particular, the SW CCRE at TOA and SRF, and the atmospheric CCRE, presented COD-dependent rates of −74 ± 4, −55 ± 5, −19 ± 2 W m−2τ−1, respectively. Additionally, increasing negative rates are found from UV to NIR for each Ci category, reflecting a higher cooling NIR contribution w.r.t. UV and VIS ranges to the SW CCRE, and being also more pronounced at the TOA w.r.t. on SRF, as expected. The contribution of the SW CCRE to the net (SW + LW) radiative balance can be also potentially relevant. These results are especially significant for space-borne photometric/radiometric instrumentation and can contribute to validation purposes of the next ESA's EarthCARE mission, whose principal scientific goal is focused on radiation-aerosol-cloud interaction research.

Cluster low-streams regression method for hyperspectral radiative transfer computations: Cases of O2 A- and CO2 bands

Cluster low-streams regression method for hyperspectral radiative transfer computations: Cases of O2 A- and CO2 bands del Águil, Ana; Efremenko, Dmitry S.; Molina García, Víctor; Kataev, Michael Yu Current atmospheric composition sensors provide a large amount of high spectral resolution data. The accurate processing of this data employs time-consuming line-by-line (LBL) radiative transfer models (RTMs). In this paper, we describe a method to accelerate hyperspectral radiative transfer models based on the clustering of the spectral radiances computed with a low-stream RTM and the regression analysis performed for the low-stream and multi-stream RTMs within each cluster. This approach, which we refer to as the Cluster Low-Streams Regression (CLSR) method, is applied for computing the radiance spectra in the O2 A-band at 760 nm and the CO2 band at 1610 nm for five atmospheric scenarios. The CLSR method is also compared with the principal component analysis (PCA)-based RTM, showing an improvement in terms of accuracy and computational performance over PCA-based RTMs. As low-stream models, the two-stream and the single-scattering RTMs are considered. We show that the error of this approach is modulated by the optical thickness of the atmosphere. Nevertheless, the CLSR method provides a performance enhancement of almost two orders of magnitude compared to the LBL model, while the error of the technique is below 0.1% for both bands.