José Agustín García

Trend analysis of monthly precipitation over the iberian peninsula for the period 1921–1995

Abstract In this study monthly and annual total precipitation records of the Iberian Peninsula are analyzed for trend. In order to take into account seasonality and serial correlation, the different months of the year are considered separately. The precipitation series were observed at forty meteorological stations scattered all over the Iberian Peninsula. The group of series were selected for their spatial and temporal representativeness. The long common period of seventy-five years (from January 1921 to December 1995) permits a reliable climatic investigation. The Wald and Wolfowitz (1943) test was applied to the series in order to detect serial correlation that could affect the subsequent trend analysis. Due to the long interval of time (one year) between consecutive data, the monthly total precipitation series for the different months showed no persistence. However, the annual total precipitation series showed a positive serial correlation. In order to detect possible trends in precipitation over the Iberian Peninsula, the Mann-Kendall test was applied to the annual and monthly series. This test is non-parametric and thus, has the advantage of being insensitive to the true (unknown) form of the distribution involved. No significant global trend was found in the annual total precipitation series. The only trend observed was a downward trend in twenty-one of the forty series of monthly total precipitation only for the month of March.

Trends in Block-Seasonal Extreme Rainfall over the Iberian Peninsula in the Second Half of the Twentieth Century

Abstract In this study trends in extreme rainfall over the Iberian Peninsula at a daily scale in the second half of the twentieth century have been detected and analyzed. For this goal 35 stations evenly distributed over the region of study covering the period 1958–97 have been studied. Two different approaches have been used. The first one consists of the nonparametric Mann–Kendall test and the Sen method. The second approach is based on the statistical theory of extreme values, involving time-dependent parameters in order to be able to reflect possible temporal changes in the frequency distribution. Results from both methods agree, confirming the reliability of the analysis. Negative trends are found for the west and southwest of the Iberian Peninsula in spring and winter. In autumn a spatial dipolar pattern appears, but trends are not so evident.

Peaks-over-Threshold Study of Trends in Extreme Rainfall over the Iberian Peninsula

A peaks-over-threshold (POT) approach is used to study trends in extreme rainfall over the Iberian Peninsula (IP) at a daily scale. Records from 52 observatories regularly distributed over Iberia with no missing data were available for the common period from 1958 to 2004. The POT approach was used because it is particularly effective at extracting information concerning true extreme events. A generalized Pareto distribution fit was made to the data involving time-dependent parameters to account for possible temporal changes in the frequency distribution. These parameters were analyzed for trends in the return-level period, of importance for engineering purposes. A time-varying threshold was defined and an automatic declustering scheme was used to select independent extreme events exceeding the threshold. The results indicate a high variability of extreme events over the coastline of the IP, greater over the Mediterranean coast than over the Atlantic coast. The calculation of the trends for the 2-yr return level yielded a large proportion of negative trends for all three seasons considered: 58% for winter, 63% for spring, and 69% for autumn. The parametric approach also revealed an increase in the area with a positive trend of the 20-yr return level relative to the 2-yr return period, especially in autumn in the east of the IP.

An Improved Outdoor Calibration Procedure for Broadband Ultraviolet Radiometers

This article aims at improving the broadband ultraviolet radiometers calibration methodology. For this goal, three broadband radiometers are calibrated against a spectrophotometer of reference. Three different one‐step calibration models are tested: ratio, first order and second order. The latter is proposed in order to adequately reproduce the high dependence on the solar zenith angle shown by the other two models and, therefore, to improve the calibration performance at high solar elevations. The proposed new second‐order model requires no additional information and, thus, keeps the operational character of the one‐step methodology. The models are compared in terms of their root mean square error and the most qualified is subsequently validated by comparing its predictions with the spectrophotometer measurements within an independent validation data subset. Results show that the best calibration is achieved by the second‐order model, with a mean bias error and mean absolute bias error lower than 2.2 and 6.7%, respectively.

Atmospheric blocking signatures in total ozone and ozone miniholes.

This paper analyzes the statistical relationship between the total ozone column (TOC) and atmospheric blocking using 40-yr European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA-40) data for the 1978-98 period, with special emphasis on winter and the European and eastern Pacific sectors. Regional blocking occurrence is accompanied by a decrease of TOC within the anticyclonic circulation region and a distinctive ozone increase upstream and downstream (upstream and south) in the Pacific (European) sector. Blocking significantly enhances the likelihood of low TOC extremes, especially over the Scandinavian and the Alaska Peninsulas, where more than 50% of winter blocks lead to TOC values in the lowest tail of the distribution. The relationship between ozone miniholes and blocking is confined to the high latitudes of both basins and is strong in Europe, where about half of the ozone miniholes occur simultaneously with blocking. Blocking-related ozone miniholes (blocking ozone miniholes) are also among the most intense and persistent. Although blocking activity does not drive the interannual variability of regional ozone miniholes, blocking ozone miniholes account for up to two-thirds of the total observed trend of ozone miniholes in Europe. The polar vortex is proposed as a feasible candidate for explaining the enhanced coupling of blocking and ozone miniholes in Europe and its long-term modulation. Blocking ozone miniholes are consistent with an almost purely dynamic origin caused by horizontal transport of ozone-poor air and vertical motions working together at different levels to reduce ozone content. Although the contribution of the former is dominant, accounting for two-thirds of ozone reduction in the 330-850-K column, the effect of the latter becomes a distinctive feature of blocking ozone miniholes.

Application of an analytical formula for UV Index reconstructions for two locations in Southwestern Spain

This paper focuses on the application of a simple analytical parameterization to the filling of the Ultraviolet Index (UVI) data gaps, and the reconstruction of past UVI values at Badajoz and Caceres (Southwestern Spain). The empirical model involves three independent variables: the solar zenith angle, the total ozone column and the clearness index. Regarding the first application, daily UVI was estimated for more than 30 days when UV measurements were not available in 2007. For these cases, the missing UVI data were replaced by estimated values, thus affecting the UVI annual mean and median. Regarding the second application, the reconstruction of past UVI time-series (1950–2000) is performed only for clear-sky cases (cloud and aerosol free conditions) using the COST 726 total ozone climatology. The linear UVI trends for two periods (1957–1978 and 1979–2000) are calculated for summer months using linear least squares fits. Both locations show statistically significant UVI trends for the most recent period 1979–2000, with values of +4.4 ± 1.6% per decade for Badajoz, and +4.9 ± 1.8% per decade for Caceres. This result is mainly driven by the ozone decline at northern mid-latitudes during this period. No significant trend is found for the other analysed period.

Empirical evaluation of a simple analytical formula for the ultraviolet index.

This paper focuses on the estimation of the UV Index (UVI) for all sky conditions using a simple analytical parameterization involving three independent variables: the solar zenith angle, the total ozone column and the clearness index. Measurements of the UVI made at Badajoz and Cáceres (Southwestern Spain) from January 2006 to December 2007 are used to estimate optimal fitting parameters for the model formula, while measurements from January to December 2008 are used to show that the formula‐based estimations have mean absolute errors lower than 6% and R2ca 0.99.

Non-stationary future return levels for extreme rainfall over Extremadura (southwestern Iberian Peninsula)

ABSTRACT Based on a previous study for temperature, a new method for the calculation of non-stationary return levels for extreme rainfall is described and applied to Extremadura, a region of southwestern Spain, using the peaks-over-threshold approach. Both all-days and rainy-days-only datasets were considered and the 20-year return levels expected in 2020 were estimated taking different trends into account: first, for all days, considering a time-dependent threshold and the trend in the scale parameter of the generalized Pareto distribution; and second, for rainy days only, considering how the mean, variance, and number of rainy days evolve. Generally, the changes in mean, variance and number of rainy days can explain the observed trends in extremes, and their extrapolation gives more robust estimations. The results point to a decrease of future return levels in 2020 for spring and winter, but an increase for autumn.

Analysis of daily rainfall processes in lower Extremadura (Spain) and homogenization of the data

SummaryIn this paper we analyze, from the point of view of stochastic processes, daily rainfall data recorded at the Badajoz Observatory (Southwestern Spain) since the beginning of the century. We attempt to identify any periodicities or trends in the daily rainfall occurrences and their dependence structure, and attempt to select the appropriate point stochastic model for the daily rainfall series.Standard regression analysis, graphical methods and the Cramer statistic show a rise in the number of cases of light rain (between 0.1 and 5 mm/d) and a decline in the number of cases of moderate to heavy rain (> 5 mm/d) in the daily rainfall at least at the 5% significance level.That the homogenization process was satisfactory is shown by the mean interarrival time of the homogenized series and the test of the rate of homogenized daily rainfall occurrences. Our analysis also shows that the behavior of the spectra of the homogenized daily rainfall counts is completely different from that of a Poisson process, so that the hypothesis of a non-homogeneous Poisson process is rejected.

Neyman-Scott cluster model for daily rainfall processes in lower extremadura (Spain): Rainfall Generating Mechanisms

SummaryA Neyman-Scott cluster model was fitted to the daily rainfall data recorded at the observatory of Badajoz (southwestern Spain) for the period 1901–1990. The data were previously homogenized. The goodness of the fit that indicated the daily rainfall process follows some Rainfall Generating Mechanism (RGM). Having decided on the criteria that a block of rainfall must fulfill to be considered as a RGM, a method was proposed to classify the days that belong to RGMs according to the 500 hPa and the surface topography. In this method each day is characterized by a string of 22 alphanumeric characters. From the subsequent analysis, the structure of the synoptic patterns associated with each RGM was deduced.