Roberto Román

Evaluation of the desert dust effects on global, direct and diffuse spectral ultraviolet irradiance

ABSTRACT This paper presents a study of a strong desert dust episode over the Iberian Peninsula, and its effect on the spectral ultraviolet (UV) irradiance in Granada, Spain. Remote sensing measurements, forecast models, and synoptic analysis are used to identify a Saharan desert dust outbreak that affected the Iberian Peninsula starting 20 July 2009. Additionally, a Bentham DMc150 spectroradiometer is employed to obtain global, direct and diffuse spectral UV irradiances every 15 minutes in Granada. The desert dust caused a large attenuation of the direct UV irradiance (up to 55%), while the diffuse UV irradiance increased up to 40% at 400 nm. The UVSPEC/LibRadtran radiative transfer model is used to study the spectral dependence of the experimental UV irradiance ratios (ratios of spectral irradiance for the day with the highest aerosol load to that measured in days with low–moderate load). The spectral increase or decrease of the UV direct irradiance ratios depends on a new parameter: a threshold wavelength. The spectral dependence of the UV diffuse irradiance ratio can be explained because under the influence of the intense dust outbreak, the Mie scattering by aerosols at shorter wavelengths is stronger than the Rayleigh scattering by gases. Finally, the sensitivity analysis of the aerosol absorption properties shows a substantial attenuation of UV spectral irradiance with a weak spectral dependence.

Uncertainty of different atmospheric ozone retrievals and its effect on temporal trends and radiative transfer simulations in the Iberian Peninsula

Uncertainty in total ozone column (TOC) values is quantified for eight different databases through a direct comparison with ground-based data at three Spanish locations, the maximum uncertainty being about 10.5 Dobson unit. A long-term TOC series is constructed using the mentioned databases from 1950 to 2011 for nine locations in the Iberian Peninsula. The monthly TOC climatology in the nine locations is presented. An exhaustive analysis is performed of TOC series trends and their statistical significance in the periods 1950–2011, 1950–1984, and 1985–2011. A significant ozone reduction between 1950 and 2011 appears in the Iberian Peninsula with a strong reduction from the late 1970s to the mid-1990s and with more or less constant or slightly increasing levels in the last 17 year. These trends are recalculated taking into account of uncertainty in the TOC values of the series, with a decrease in the number of statistically significant trends emerging. The statistically significant trends in annual and in the averaged Iberian Peninsula series are usually still significant even considering the uncertainty. Finally, a study is carried out of the uncertainty caused by TOC uncertainty in total shortwave and erythemal ultraviolet irradiances simulated under cloudless skies using a radiative transfer model.

Comparison of total water vapor column from GOME-2 on MetOp-A against ground-based GPS measurements at the Iberian Peninsula

Water vapor column (WVC) obtained by GOME-2 instrument (GDP-4.6 version) onboard MetOp-A satellite platform is compared against reference WVC values derived from GPS (Global Positioning System) instruments from 2007 to 2012 at 21 places located at Iberian Peninsula. The accuracy and precision of GOME-2 to estimate the WVC is studied for different Iberian Peninsula zones using the mean (MBE) and the standard deviation (SD) of the GOME-2 and GPS differences. A direct comparison of all available data shows an overestimation of GOME-2 compared to GPS with a MBE of 0.7 mm (10%) and a precision quantified by a SD equals to 4.4mm (31%). South-Western zone presents the highest overestimation with a MBE of 1.9 mm (17%), while Continental zone shows the lowest SD absolute value (3.3mm) due mainly to the low WVC values reached at this zone. The influence of solar zenith angle (SZA), cloud fraction (CF), and the type of surface and its albedo on the differences between GOME-2 and GPS is analyzed in detail. MBE and SD increase when SZA increases, but MBE decreases (taking negative values) when CF increases and SD shows no significant dependence on CF. Under cloud-free conditions, the differences between WVC from GOME-2 and GPS are within the WVC error given by GOME-2. The changes of MBE and SD on Surface Albedo are not so evident, but MBE slightly decreases when the Surface Albedo increases. WVC from GOME-2 is, in general, more precise for land than for sea pixels.

Validation of MODIS integrated water vapor product against reference GPS data at the Iberian Peninsula

Abstract In this work, the water vapor product from MODIS (MODerate-resolution Imaging Spectroradiometer) instrument, on-board Aqua and Terra satellites, is compared against GPS water vapor data from 21 stations in the Iberian Peninsula as reference. GPS water vapor data is obtained from ground-based receiver stations which measure the delay caused by water vapor in the GPS microwave signals. The study period extends from 2007 until 2012. Regression analysis in every GPS station show that MODIS overestimates low integrated water vapor (IWV) data and tends to underestimate high IWV data. R 2 shows a fair agreement, between 0.38 and 0.71. Inter-quartile range (IQR) in every station is around 30–45%. The dependence on several parameters was also analyzed. IWV dependence showed that low IWV are highly overestimated by MODIS, with high IQR (low precision), sharply decreasing as IWV increases. Regarding dependence on solar zenith angle (SZA), performance of MODIS IWV data decreases between 50° and 90°, while night-time MODIS data (infrared) are quite stable. The seasonal cycles of IWV and SZA cause a seasonal dependence on MODIS performance. In summer and winter, MODIS IWV tends to overestimate the reference IWV value, while in spring and autumn the tendency is to underestimate. Low IWV from coastal stations is highly overestimated (∼60%) and quite imprecise (IQR around 60%). On the contrary, high IWV data show very little dependence along seasons. Cloud-fraction (CF) dependence was also studied, showing that clouds display a negligible impact on IWV over/underestimation. However, IQR increases with CF, except in night-time satellite values, which are quite stable.

A method to determine the ozone radiative forcing in the ultraviolet range from experimental data

This paper proposes a method to calculate the ozone radiative forcing (RF) at surface in the ultraviolet (UV) spectral range for all-sky conditions based on the estimation of the ozone efficiency (OE) from experimental data that are subsequently applied to changes on the total ozone column (TOC) since late 1970s. The OE is defined as the rate at which the solar UV irradiance is “forced” per TOC unit, being estimated for all-sky conditions from UV-B (280–320 nm) and TOC data recorded with a Bentham spectroradiometer at Granada (Spain). The results showed a clear seasonal pattern in the OE values with largest monthly averages (in absolute terms) in July (−4.2 ± 0.3 mW/m2 per Dobson Unit) and the smallest in January (−0.7 ± 0.3 mW/m2 per Dobson Unit). The continuous and consistent TOC data set (1979–2008) provided by the Multisensor Reanalysis over the study site showed that spring months present the largest annual TOC changes relative to 1979 while summer months exhibit small variations. Thus, spring has the largest contribution (~53%) to annual ozone RF followed by summer (~17%), winter (~16%), and autumn (~4%). The evolution of the ozone RF relative to 1979 in the UV-B range at Granada showed positive values for most of years (between 5 and 40 mW/m2). Finally, the long-term evolution of the ozone RF exhibited a positive trend until the mid-1990s and, subsequently, a weak negative trend until the end of the analyzed period.

Comparison of integrated water vapor from GNSS and radiosounding at four GRUAN stations

Integrated water vapor (IWV) data from Global Navigation Satellite Systems (GNSS) and radiosounding (RS) are compared over four sites (Lindenberg, Ny-Ålesund, Lauder and Sodankylä), which are part of the Global Climate Observing System (GCOS) Reference Upper Air Network (GRUAN). Both datasets show an excellent agreement, with a high degree of correlation (R2 over 0.98). Dependences of GNSS-RS differences on several variables are studied in detail. Mean bias error (MBE) and standard deviation (SD) increase with IWV, but in relative term, these variables decrease as IWV increases. The dependence on solar zenith angle (SZA) is partially related to the distribution of IWV with SZA, but the increase of SD for low SZA could be associated with errors in the humidity sensor. Large surface pressures worsen performance, which could be due to the fact that low IWV is typically present in high pressure situations. Cloud cover shows a weak influence on the mentioned MBE and SD. The horizontal displacement of radiosondes generally causes SD to increase and MBE to decrease (increase without sign), as it could be expected. The results point out that GNSS measurements are useful to analyze performance to other instruments measuring IWV.

Assessment of Sun photometer Langley calibration at the high-elevation sites Mauna Loa and Izaña

The aim of this paper is to analyze the suitability of the high-mountain stations Mauna Loa and Izaña for Langley plot calibration of Sun photometers. Thus the aerosol optical depth (AOD) characteristics and seasonality, as well as the cloudiness, have been investigated in order to provide a robust estimation of the calibration accuracy, as well as the number of days that are suitable for Langley calibrations. The data used for the investigations belong to AERONET and GAW-PFR networks, which maintain reference Sun photometers at these stations with long measurement records: 22 years at Mauna Loa 5 and 15 years at Izaña. In terms of clear sky and stable aerosol conditions, Mauna Loa (3397m a.s.l.) exhibits on average of 377 Langleys (243 morning and 134 afternoon) per year suitable for Langley plot calibration, whereas Izaña (2373m a.s.l.) shows 343 Langleys (187 morning and 155 afternoon) per year. The background AOD(500nm) values, on days that are favorable for Langley calibrations, are in the range 0.01-0.02 throughout the year, with well defined seasonality that exhibits a spring maximum at both stations plus a slight summer increase at Izaña. The statistical analysis of the long-term determination of 10 extraterrestrial signals yields to a calibration uncertainty of ~0.2-0.5%, being this uncertainty smaller in the near infrared and larger in the ultraviolet wavelengths. This is due to atmospheric variability that cannot be reduced based only on quality criteria of individual Langely plots.