Daily Precipitation Sums Research Articles

Interannual variations in the seasonal cycle of extreme precipitation in Germany and the response to climate change

Abstract. Annual maxima of daily precipitation sums can be typically described well with a stationary generalized extreme value (GEV) distribution. In many regions of the world, such a description does also work well for monthly maxima for a given month of the year. However, the description of seasonal and interannual variations requires the use of non-stationary models. Therefore, in this paper we propose a non-stationary modeling strategy applied to long time series from rain gauges in Germany. Seasonal variations in the GEV parameters are modeled with a series of harmonic functions and interannual variations with higher-order orthogonal polynomials. By including interactions between the terms, we allow for the seasonal cycle to change with time. Frequently, the shape parameter ξ of the GEV is estimated as a constant value also in otherwise instationary models. Here, we allow for seasonal–interannual variations and find that this is beneficial. A suitable model for each time series is selected with a stepwise forward regression method using the Bayesian information criterion (BIC). A cross-validated verification with the quantile skill score (QSS) and its decomposition reveals a performance gain of seasonally–interannually varying return levels with respect to a model allowing for seasonal variations only. Some evidence can be found that the impact of climate change on extreme precipitation in Germany can be detected, whereas changes are regionally very different. In general, an increase in return levels is more prevalent than a decrease. The median of the extreme precipitation distribution (2-year return level) generally increases during spring and autumn and is shifted to later times in the year; heavy precipitation (100-year return level) rises mainly in summer and occurs earlier in the year.

Robust weather-adaptive post-processing using model output statistics random forests

Abstract. Physical numerical weather prediction models have biases and miscalibrations that can depend on the weather situation, which makes it difficult to post-process them effectively using the traditional model output statistics (MOS) framework based on parametric regression models. Consequently, much recent work has focused on using flexible machine learning methods that are able to take additional weather-related predictors into account during post-processing beyond the forecast of the variable of interest only. Some of these methods have achieved impressive results, but they typically require significantly more training data than traditional MOS and are less straightforward to implement and interpret. We propose MOS random forests, a new post-processing method that avoids these problems by fusing traditional MOS with a powerful machine learning method called random forests to estimate weather-adapted MOS coefficients from a set of predictors. Since the assumed parametric base model contains valuable prior knowledge, much smaller training data sizes are required to obtain skillful forecasts, and model results are easy to interpret. MOS random forests are straightforward to implement and typically work well, even with no or very little hyperparameter tuning. For the difficult task of post-processing daily precipitation sums in complex terrain, they outperform reference machine learning methods at most of the stations considered. Additionally, the method is highly robust in relation to changes in data size and works well even when less than 100 observations are available for training.

Regional variation of climatological cloudburst frequency estimated from historical observations of daily precipitation sums

AbstractThis study aims to estimate climatological cloudburst frequencies from daily precipitation sums for a dense network of 161 historical Danish stations covering the period 1914–2010, in order to reveal any systematic regional differences. We define a cloudburst day for a given station as a day when the maximum 1‐h precipitation sum exceeds a predefined threshold; we use the thresholds 15, 20 and 25 mm, corresponding to a cloudburst frequency around 0.42, 0.16 and 0.06 year−1, respectively. Supplementing the historical data, we use hourly data from a modern network and relate the daily probability of cloudburst occurrence to the corresponding daily precipitation using binary regression. This allows an estimation of the cloudburst frequency from a series of daily sums. To validate the methods, we use 33 stations that have been operating for 30 years or longer. For these stations, we demonstrate significant skill by comparing observed and estimated cloudburst frequencies with a jackknife procedure. By applying the model to the 161 historical series, we estimate climatological cloudburst frequencies throughout Denmark. We find large systematic regional variations across Denmark and these variations are much larger, but with a similar regional pattern, than obtained by previous work. We suggest applying the approach to the historical daily precipitation series that exist across the world.

Data quality control and homogenization of daily precipitation and air temperature (mean, max and min) time series of Ukraine

AbstractIn this paper we present the results of quality control and homogenization procedures applied to long time series of daily atmospheric precipitation sums (Rr) and daily mean (Tm), maximum (Tx) and minimum (Tn) air temperature collected in Ukraine. The daily data from 178 meteorological stations covering the period of 1946–2020 were analysed. In order to perform a thorough quality assurance check, we used the R package INQC, while the Climatol homogenization software was used to detect and remove breaks from the time series. The INQC quality assurance tests revealed a relatively small number of erroneous records (around 0.01% for each variable) and suspicious values (up to 0.09%). The application of Climatol resulted in 195, 296, 355 and 359 break points, detected for Rr, Tm, Tx and Tn, respectively. These quantities coincide roughly with the results of the HOMER homogenization procedure applied to monthly time series for the same stations and almost the same period (performed in the previous works of the authors). To verify the homogenization results, statistical comparison of the raw and homogenized time series was performed. The verification demonstrated that the quality control and homogenization procedures detected and removed errors and breaks very well, and air temperature and precipitation fields after the homogenization are more self‐consistent compared to the original raw data.

Recurrence of Drought Events Over Iberia. Part II: Future Changes Using Regional Climate Projections

Seasonal droughts are a common feature of the Iberian (Mediterranean) climate. They can have severe impacts on both natural and human life – especially, when recurring in consecutive years. In this study, we investigate the potential impacts of climate change on recurrent drought events in the Iberian Peninsula (IP). With this aim, we use the new set of indices introduced in Moemken and Pinto (2022): the Recurrent Dry Year Index (RDYI) and the Consecutive Drought Year (CDY) Index. These are applied to a large EURO-CORDEX multi-model ensemble consisting of 25 different global-to-regional model (GCM-RCM) chains that follow the RCP8.5 scenario with 12 km horizontal resolution. A drizzle correction and a simple multiplicative approach are used to bias-adjust the daily precipitation sums. Results reveal a general tendency towards more severe drought conditions in IP under different global warming levels (GWLs). Moreover, recurrent drought events are projected to occur more frequent and last longer. While the ensemble mean responses are only moderate for a GWL of +2°C (compared to the pre-industrial average), recurrent drought conditions are strongly enhanced for the +3°C GWL. The magnitude of projected changes shows some sensitivity on the choice of index and model. Typically, changes are more pronounced for indices based on the effective drought index (EDI) and show a larger spread for the individual GCMs than for the various RCMs. Nevertheless, the climate change signals are robust for most of IP and all indices, with a larger model agreement for the +3°C GWL. We conclude that the Iberian Peninsula is confronted with an increased risk of recurrent drought events in future decades. If global warming should exceed the +3°C threshold, the majority of models projects an almost permanent state of drought – which could result in severe implications for the Iberian population and ecosystems.

NEX-GDDP Multimodel Ensemble vs. E-OBS—Evaluation of the Extreme Temperatures and Precipitation over Southeast Europe: Historical Comparison

The present study evaluates the National Aeronautics Space Administration (NASA) Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) dataset that provides statistically downscaled CMIP5 historical and future climate projections of the daily precipitation sum and extreme temperatures at high spatial resolution. A multimodel ensemble from all 21 available models is composed and compared against gridded observations from E-OBS. The study is performed over Southeast Europe for the whole time span of the historical period of NEX-GDDP 1950–2005. The performance of the NEX-GDDP data was evaluated at multiple time scales such as annual, seasonal, monthly, and daily. The skill of the multimodel ensemble to reproduce the interannual variability, as well as the long-term trend, is also evaluated. Moreover, key climate indices of the Expert Team on Climate Change Detection and Indices (ETCCDI), derived from the ensemble extreme temperatures and precipitation are superimposed on their counterparts based on the reference dataset E-OBS. Findings of the performed research indicate that NEX-GDDP parameters are in good agreement with the reference over the considered period on monthly, seasonal and annual scales which agrees with the outcomes from similar studies for other parts of the world. There are also no systematic differences in the pattern of the biases of the minimum and maximum temperature. Generally, the multimodel ensemble reproduces the extreme temperatures significantly better than the precipitation sum. The analysis reveals also the nonnegligible inefficiency of the NEX-GDDP ensemble to reproduce the long-term trend of the considered parameters as well as the climate extremes expressed with the ETCCDI indices.

Creation of a representative climatological database for Hungary from 1870 to 2020

Climate studies, particularly those that are related to climate change, require long, high-quality controlled data sets, which are representative both spatially and temporally. Changing the conditions of measurements, for example relocating the station, or changing the frequency and timing of measurements, or changing the instruments used can cause breaks in the time series. To avoid these problems, data errors and inhomogeneities are eliminated and the data gaps are filled by using the MASH (Multiple Analysis of Series for Homogenization, Szentimrey, 1999, 2008) homogenization procedure. The Hungarian meteorological observation network was upgraded significantly in the last decades. Homogenization of the data series raises the question of how to homogenize long and short data series together within the same process. It is possible to solve this with the MASH method due it has solid mathematical foundations, which make it suitable for such purposes. The solution includes the synchronization of the common parts’ inhomogeneities within three (or more) different MASH processing of the three (or more) datasets with different lengths depending on the time periods and elements. After the homogenization process, the station data series were interpolated to a 0.1 degree regular grid covering the whole area of Hungary. The MISH (Meteorological Interpolation based on Surface Homogenized Data Basis; Szentimrey and Bihari, 2007) program system was used for this purpose. The MISH procedure was developed specifically for the interpolation of various meteorological elements. In the case of mean temperature, we also renewed the MISH modeling, as compared to previous years, the number of homogenized stations doubled due to the new work, so it was expedient to model the climate statistical parameters with this extended station system. Time series of daily mean temperature and precipitation sum for the period 1870–2020 for Hungary were used in this study. As a result, the longest ever homogenized, gridded daily data sets became available for Hungary. The method described here can also be applied to produce representative datasets for other meteorological elements.

Technical Note: Temporal disaggregation of spatial rainfall fields with generative adversarial networks

Abstract. Creating spatially coherent rainfall patterns with high temporal resolution from data with lower temporal resolution is necessary in many geoscientific applications. From a statistical perspective, this presents a high- dimensional, highly underdetermined problem. Recent advances in machine learning provide methods for learning such probability distributions. We test the usage of generative adversarial networks (GANs) for estimating the full probability distribution of spatial rainfall patterns with high temporal resolution, conditioned on a field of lower temporal resolution. The GAN is trained on rainfall radar data with hourly resolution. Given a new field of daily precipitation sums, it can sample scenarios of spatiotemporal patterns with sub-daily resolution. While the generated patterns do not perfectly reproduce the statistics of observations, they are visually hardly distinguishable from real patterns. Limitations that we found are that providing additional input (such as geographical information) to the GAN surprisingly leads to worse results, showing that it is not trivial to increase the amount of used input information. Additionally, while in principle the GAN should learn the probability distribution in itself, we still needed expert judgment to determine at which point the training should stop, because longer training leads to worse results.

Numerical simulation of intense precipitation in Moscow region: a case study of a heavy rainfall event on June 30, 2017

This paper considers physical and synoptic mechanisms of an extreme rainfall on June 30, 2017 in the central part of European Russia, which was the maximum of daily precipitation sum in Moscow (65 mm) since 1970. Based on meteorological observations, radar data, and ERA5 reanalysis data, we show that the rainfall was associated with three mesoscale convective systems (MCSs) that formed in the warm sector of a cyclone in a strip of anomalously high moisture content in the considered region, which developed further due to advection and evaporation. A numerical simulation with mesoscale model COSMO shows a significant contribution to the precipitation intensity of evaporation from the Earth’s surface: a 10-times decrease in soil the moisture in the initial conditions leads to a 3-times decrease in the amount of precipitation and its intensity. Additionally, we consider urban-induced effects on this rainfall event by switching on and off urban parametrization TERRA_URB. The application of the urban surface parametrization has not changed the average amount of precipitation in Moscow region. It causes, however, a redistribution of precipitation sums within it.

Determination of N-year Design Precipitation in the Czech Republic by Annual Maximum Series Method

The sum of design precipitation of a selected repetition period, provided that it is evenly distributed over the river basin area, is a basic input for the calculation of the direct outflow volume by the curve number method. It is necessary to determine the design precipitation for each location using the statistical methods and the longest available data series on daily precipitation sums, or more specifically their annual maximums. This paper deals with the determination of design precipitation from data of eight stations of the Czech Hydrometeorological Institute for the period 1961–2013. From a series of annual maximum values of daily precipitation sums, N‑year design precipitations were calculated using two methods (Gumbel and generalized extreme value distributions). The conformity of both models with empirical distribution of values was statistically tested to evaluate which of the models gave more accurate results. In these cases, it was more appropriate to use the generalized extreme value distribution. Finally, the newly calculated characteristics were compared with the design values used by Šamaj et al. (1985), where significant differences were found.

Використання інформації відкритих баз метеоданих у дослідженнях новітніх кліматичних змін в західному регіоні України

Studies of the regional manifestations of recent climate changes have a big practical importance for different scientific fields and applied purposes. Open databases of historical weather conditions can serve as a valuable source of information on recent climate changes for Ukrainian scientists with limited funds available. These data can be preprocessed, analyzed and plotted with open-source data processing software. ECAD and GHCN datasets have a plenty of uploadable weather data that can be processed with the free R software environment. ECAD has data for 39 Ukrainian weather stations, while GHCN database contains temperature and precipitation data for 190 Ukrainian weather stations, although with variable time span and completeness. Preprocessing of these data in R may involve creating special scripts to automatize and speed up the work. The data on daily temperatures and precipitation sums for three weather stations in Western Ukraine (Lviv, Chernivtsi, and Uzhhorod) were uploaded from GHCN dataset for the common period of 1960 – 2014 yrs. The completeness of temperature data was significantly higher (less than 0.16% dates omitted) than that of precipitation data (11 – 12% dates omitted). The date field was split to three separate fields for year, month and day, and weather data were summarized to obtain yearly average temperatures and precipitation sums. In addition, separate temperature averages and precipitation sums were calculated for each of four seasons by filtering data with appropriate month numbers. Linear and quadratic models were built on derived time series to detect possible trends. Coefficients of determination were calculated and analyzed for significance for linear and quadratic models. For significant linear trends, regression coefficients were calculated which characterize the rate of change average for the time period, while in case of significant quadratic trends theirs directions were determined. It was established that while average annual temperatures have significantly increased in each of the three analyzed locations, this increase was nearly twice as large in Chernivtsi than that in Uzhhorod. In Lviv, temperatures have increased the most in spring and winter, while in other two locations the summer increase was the largest. Summer temperatures were increasing with acceleration, which can be a warning sign. Autumn was the only season when temperatures did not increase significantly in any of the analyzed locations. Trends in precipitation amounts were much less salient: precipitation somewhat decreased in Chernivtsi in winter and spring, and somewhat increased in Lviv in spring.

Assessing the Use of Satellite-Based Estimates and High-Resolution Precipitation Datasets for the Study of Extreme Precipitation Events over the Iberian Peninsula

An assessment of daily accumulated precipitation during extreme precipitation events (EPEs) occurring over the period 2000–2008 in the Iberian Peninsula (IP) is presented. Different sources for precipitation data, namely ERA-Interim and ERA5 reanalysis by the European Centre for Medium-Range Weather Forecast (ECMWF) and Tropical Rainfall Measuring Mission (TRMM) Multisatellite Precipitation Analysis (TMPA), both in near-real-time and post-real-time releases, are compared with the best ground-based high-resolution (0.2° × 0.2°) gridded precipitation dataset available for the IP (IB02). In this study, accuracy metrics are analysed for different quartiles of daily precipitation amounts, and additional insights are provided for a subset of EPEs extracted from an objective ranking of extreme precipitation during the extended winter period (October to March) over the IP. Results show that both reanalysis and multi-satellite datasets overestimate (underestimate) daily precipitation sums for the least (most) extreme events over the IP. In addition, it is shown that the TRMM TMPA precipitation estimates from the near-real-time product may be considered for EPEs assessment over these latitudes. Finally, it is found that the new ERA5 reanalysis accounts for large improvements over ERA-Interim and it also outperforms the satellite-based datasets.

Modeling Water and Heat Balance Components for Large Agricultural Region Utilizing Information from Meteorological Satellites

The method has been developed to evaluate water and heat balance components for vegetation covered area of regional scale based on the refined physical-mathematical model of vertical water and heat exchange between land surface and atmosphere (Land Surface Model, LSM) for vegetation season adapted to satellite information on land surface and meteorological conditions. The LSM is accommodated for utilizing satellite-derived estimates of vegetation and meteorological characteristics as model parameters and input variables. Estimates of these characteristics presented as distributions of their values over the study area have been obtained from AVHRR/NOAA, MODIS/EOS Terra and Aqua, SEVIRI/Meteosat-9, -10 data. To build such estimates methods and technologies have been developed and refined using results of thematic processing measurement data from these sensors. Among them the original Multi Threshold Method (MTM) has been developed and tested to calculate daily precipitation sums using rainfall intensity estimates retrieved from AVHRR and SEVIRI data with subsequent replacement of ground-measured rainfall amounts by these daily rainfalls. All technologies have been adapted to the study area with square of 227300 km2 being the part of the Central Black Earth Region of European Russia. Developed earlier procedures of utilizing satellitederived estimates of vegetation and meteorological characteristics (including precipitation) in the model have been refined and verified. Final result of modeling is the fields of soil water content, evapotranspiration and other water and heat balance components of the region under study for years 2012–2014 vegetation seasons.

Evaluation of dry matter accumulation of maize (Zea mays L.) hybrids

The increase of the grain yield of maize is closely correlated with its seasonal dry matter accumulation. Dry matter is accumulated into the grain yield during the grain filling period. The following maize hybrids were involved in the experiment: Armagnac FAO 490, Loupiac FAO 380 and Sushi FAO 340. In order to determine dry matter content, two samples per week were taken on the following days: 22nd, 25th, 28th, 31st August, 4th, 7th, 14th, 18th, 22nd, 25th, 29th September and 2nd, 6th, 9th, 13th October. In the course of sampling the weight of 100 grains from the middle section of 4 ears was measured in 4 replications. Dry matter content was determined after drying to constant weight in a drying cabinet at 60 °C. Harvesting was performed on 13th October 2017.
 The daily precipitation sum was determined by local measurements, while the daily radiation and temperature data were provided by the Meteorological Observatory Debrecen of the National Meteorological Service in Budapest. Among the agrometeorological parameters, an analysis was made of the precipitation during the growing season, effective heat sums during the vegetative and generative phase, and the water supplies. The daily heat sums were determined using the algorithm proposed.
 The amount of precipitation in the winter period before the 2017 growing season was 210 mm. The soil was saturated until its field capacity. The rather dry and warm March and April had a favourable effect, but there was no worthy amount of precipitation until May (51 mm) due to the condition of the dried seedbed. Sowing was performed on the 5th of May 2017 in a randomised small plot experiment. There was favourable precipitation and temperature during the growing season, thereby providing ideal conditions for maize development, growth and yield formation. There was near average amount of precipitation in each year. The total amount of precipitation in the summer period is 342 mm. Temperature was mostly above the average, but there was no long and extremely warm period.
 The Armagnac hybrid reached its highest dry matter mass 126 days after emergence. Physiological maturity was reached sooner (on the 119th day) in the case of Loupiac, and even sooner in the case of Sushi (116th day). The thousand grain weight of Sushi (which has the shortest ripening period) was 286 g at the time of physiological maturity, while that of Loupiac was 311 g. Compared to Sushi, Armagnac showed 12 g more dry matter accumulation (306 g). In the case of all three examined hybrids, physiological maturity was preceded by an intensive phase, when the dynamics of dry matter accumulation was rather quick. On average, Sushi gained 2.8 g dry matter per day between 103 days following emergence and physiological maturity, while the same values were 3.2 g for Armagnac and 3.3 g for Loupiac. The aim of the regression line slope is to predict the behavior of the dependent variable with the knowledge of the values and characteristics of the independent variables using the regression line equation. Furthermore, to determine how the location affected the dynamic of dry matter accumulation in the Armagnac, Loupiac and Sushi hybrids. In regression analysis, the coefficient of explanation showed that the effect of day in the Armagnac was 97%, in the Loupiac 94%, in the Sushi 90 %. The determination coefficient (R2) is useful in determing how the regression equation fits. But, as we have seen, the determination coefficient alone is not sufficient to verify the model’s accuracy, in addition to the determination coefficient (R2), the normality of the data or the residuals, the variance of the variables at different levels, the independence of the data relative to time and non-oblique. Observations are evaluated for the correctness of the fitted model.
 Dry matter values decreased evenly and slightly following physiological maturity. According to our research results, it was established that physiological maturity is followed by a moderate dry matter loss. Until harvesting, Armagnac lost 40 g of its thousand mass weight in 29 days, while the same value pairs were 69 g in 36 days for Loupiac and 29 g in 39 days for Sushi. Loupiac – which had the highest weight at the time of physiological maturity – lost the most of its dry weight; therefore, Armagnac and Sushi had higher values at the time of harvesting.

Seasonal Cycle in German Daily Precipitation Extremes

The seasonal cycle of extreme precipitation in Germany is investigated by fitting statistical models to monthly maxima of daily precipitation sums for 2,865 rain gauges. The basis is a non-stationary generalized extreme value (GEV) distribution variation of location and scale parameters. The negative log-likelihood serves as the forecast error for a cross validation to select adequate orders of the harmonic functions for each station. For nearly all gauges considered, the seasonal model is more appropriate to estimate return levels on a monthly scale than a stationary GEV used for individual months. The 100-year return-levels show the influence of cyclones in the western, and convective events in the eastern part of Germany. In addition to resolving the seasonality, we use a simulation study to show that annual return levels can be estimated more precisely from a monthly-resolved seasonal model than from a stationary model based on annual maxima.

A spatial and seasonal climatology of extreme precipitation return-levels: A case study

A spatial and seasonal modeling approach for precipitation extremes is introduced and exemplified for the Berlin–Brandenburg region in Germany. Monthly maxima of daily precipitation sums are described with a generalized extreme value distribution (GEV) with spatially and seasonally varying parameters. This allows for a return-level prediction also at ungauged sites. The seasonality is captured with harmonic functions, spatial variations are modeled with Legendre polynomials for longitude, latitude and altitude. Interactions between season and space allow for a spatially varying seasonal cycle. Orders of the harmonic and Legendre series are determined using a step-wise forward regression approach with the Bayesian Information Criterion (BIC) as model selection criterion. The longest 80 series are used to verify the approach in a cross-validation experiment based on the Quantile Skill Score (QSS). The model presented describes the observations at all these stations more accurately than a GEV applied to each month and location separately. These improvements are due to the assumption of smoothly varying GEV parameters in time and space; information from neighboring observations in time and space are used to obtain parameters at a given location. Apart from robustness, this approach allows also a seasonally and spatially varying shape parameter and results are found to be more accurate.

Climate-growth analysis using long-term daily-resolved station records with focus on the effect of heavy precipitation events

Oak tree-ring series contain detailed information regarding climate variability for centuries to millennia. Such reconstructions are compiled with the use of climate response functions that are typically based on monthly precipitation or temperature data sets. We present an approach using a MATLAB® script and long-term daily precipitation and daily mean temperature records to evaluate intervals with daily resolution of radial growth sensitivity and to determine the effect of heavy precipitation events in the Mainfranken region (southern Germany). This allows improved insights into tree-ring response for local climate reconstructions. Annual radial stem growth is highly sensitive to total cumulative precipitation during the current year spring-summer period, but less sensitive to daily mean temperature. Response analysis reveals better results to precipitation records, when (very) heavy precipitation events are omitted (April 14–July 18, R2=0.31, p<0.01). Temporal sensitivity analysis of total ring width (TRW) to precipitation response within nine sub-periods revealed that the length of the sensitive intervals stretch between 41 to 141 days, depending on the period investigated. Our study shows that annual radial growth of oak trees is mainly affected by daily precipitation sums (DPS) of less than 10mm. In contrast, heavy rainfall events do not influence radial increment significantly, but may substantially increase the total precipitation sum of the growing season. We propose that oaks in the Mainfranken region contain low to moderate information in their tree rings regarding heavy precipitation events. Furthermore, we conclude that a disordered sensitivity of TRW to precipitation during the past three to four decades is caused by drought climatologies, increased sulfur dioxide emission in the study area as well as by changes in diurnal temperature range. Thus, climate response functions for hydroclimatic reconstructions may only be developed until the 1970s.

Ensemble Postprocessing of Daily Precipitation Sums over Complex Terrain Using Censored High-Resolution Standardized Anomalies

Probabilistic forecasts provided by numerical ensemble prediction systems have systematic errors and are typically underdispersive. This is especially true over complex topography with extensive terrain-induced small-scale effects, which cannot be resolved by the ensemble system. To alleviate these errors, statistical postprocessing methods are often applied to calibrate the forecasts. This article presents a new full-distributional spatial postprocessing method for daily precipitation sums based on the standardized anomaly model output statistics (SAMOS) approach. Observations and forecasts are transformed into standardized anomalies by subtracting the long-term climatological mean and dividing by the climatological standard deviation. This removes all site-specific characteristics from the data and makes it possible to fit one single regression model for all stations at once. As the model does not depend on the station locations, it directly allows the creation of probabilistic forecasts for any arbitrary location. SAMOS uses a left-censored power-transformed logistic response distribution to account for the large fraction of zero observations (dry days), the limitation to nonnegative values, and the positive skewness of the data. ECMWF reforecasts are used for model training and to correct the ECMWF ensemble forecasts with the big advantage that SAMOS does not require an extensive archive of past ensemble forecasts as only the most recent four reforecasts are needed, and it automatically adapts to changes in the ECMWF ensemble model. The application of the new method to the central Alps shows that the new method is able to depict the small-scale properties and returns accurate fully probabilistic spatial forecasts.

The role of atmospheric circulation patterns in agroclimate variability in finland, 1961–2011

This study evaluates interannual variations and trends in growing season daily temperature sum and daily precipitation sum in Finland during 1961–2011, and their connections to well known atmospheric circulation patterns. Changes in summer (June–August) climate partially explain changes in growing season daily temperature sum and daily precipitation sum over Finland, which naturally decreased from south to north. On a national scale, growing season warmed and became wetter during 1961–2011, as growing season daily temperature sum and daily precipitation sum significantly (p < 0.05) increased by 5.01 ± 3.17°C year–1 and 1.39 ± 0.91 mm year–1, respectively. The East Atlantic pattern was the most influential atmospheric circulation pattern for variations in growing season daily temperature sum (rho = 0.40) across Finland and the East Atlantic/West Russia pattern was most influential for growing season daily precipitation sum variability (rho = –0.54). There were significant (p < 0.05) increasing trends in growing season daily temperature sum and daily precipitation sum throughout Finland during 1961–2011. Increased growing season daily temperature sum was mainly observed in northern, central, western, eastern and coastal areas of south‐western Finland. This warming was positively associated with the East Atlantic pattern in the north, centre and south, but negatively associated with the East Atlantic/West Russia pattern in eastern Finland. Increased GSP mostly occurred in southern, eastern, western, central, northern and north‐western Finland. These wetting trends were positively correlated with the East Atlantic pattern in the north and negatively correlated with the Polar pattern in the south and the East Atlantic/West Russia pattern in the east, west, centre and north‐east of Finland. The overall agroclimatic year‐to‐year variability in Finland between 1961 and 2011 was mostly linked to variations in the East Atlantic and East Atlantic/West Russia patterns.

Heavy rains in Brno Region (The Czech Republic)

This study evaluates occurrence of the heavy rains at urban areas on the example of Brno agglomeration. Extreme precipitation events are studied with the help of 1-minute records from three urban sites in the Brno and one site in the rural region. Additionally data from nine Hellmann rain gauges from Brno and surroundings were used. The event of torrential rains was evaluated according to Wussow’s method. The study found that torrential rains occurred in Brno in the season from April to November. The fifteen-year long observation of the phenomenon at Brno-Žabovřesky station shows varying frequency of torrential rainfall - from none in 2007 to four in 2004 and 2010. The most extremal was catastrophic downpour on July 15, 2009 when precipitation sum reached total up to 42.7 mm in 52 minutes and the maximal rainfall intensity was 4.1 mm/min. This phenomenon occurs only one day after downpour on July 14, 2009. The overall number of torrential episodes is not very high, but it is compensated by the fact that they can repeat in a relatively short period of time, due to specific synoptic conditions. Maximum rainfall intensity during torrential rains reaches values of about 3 mm / min and the average proportion of torrential rain in the daily precipitation sum is around 70%. The phenomenon of heavy rains represents a significant problem for densely populated urban areas because such extreme rainfall has a large impact on the hydrological system of cities.