Radar‐based summer precipitation climatology of the Czech Republic

Abstract

To assess the climatology of the Czech Republic (CR) with a high spatial (1 km) resolution, this study uses radar-based precipitation data collected over the summer seasons of a 10-year period (2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011). Radar reflectivity data were obtained from two C-band Doppler weather radars, integrated in time and merged with daily precipitation totals from rain gauge measurements. Using radar measurements, daily adjusted precipitation totals were later divided into 10-min precipitation totals that were continuously accumulated in a series of time windows ranging from 30 min to 24 h. This climatological analysis confirms that altitude substantially influences not only the spatial distribution of mean precipitation but also the spatial distribution of precipitation maxima as well as the diurnal cycle of precipitation in the CR. While the 24-h totals are generally maximal in mountains, mean and absolute seasonal maxima of short-term totals (up to 1 and 6 h, respectively) are detected in altitudes between 300 and 600 m a.s.l. Regarding the diurnal cycle, maximum frequency of precipitation occurs 2 h earlier in the mountains, whereas mean totals remain at the same level until 2100 UTC. The mean time during which precipitation maxima occur generally does not change with altitude. Nevertheless, a detailed regional study demonstrates that short-term precipitation maxima usually start earlier in the afternoon in and around mountainous regions. Long-term (mainly 6-h) precipitation maxima occur later than short ones but are substantially less concentrated in time, especially in the mountains. These differences between mountains and lowlands can be explained by smaller relative proportions and earlier onset of convective precipitation in mountains.

Characterising patterns of heavy precipitation events in the eastern Mediterranean using a weather radar and convection-permitting WRF simulations

“…rain maps in which at least one convective rain cell, defined as a connected region 3 km with rain intensity exceeding 10 mm h -1 and including at least one pixel exceeding 25 mm h -1 , is observed (Marra and Morin, 2018). We computed the 2-D spatial autocorrelation function of the convective fields following the method in Nerini et al (2017). A three-parameter exponential function (Eq.…”

“…Remotely-sensed precipitation estimates, such as those acquired from weather radars, provide the necessary spatiotemporal resolutions (e.g., 1 km, 5 min) and coverage (regional scale), and have been shown to be useful for analysing specific events (e.g., Borga et al, 2007;Dayan et al, 2001;Krichak et al, 2000;Smith et al, 2001). Where continuous radar records exist, they have been used in climatological studies as well (Belachsen et al, 2017;Bližňák et al, 2018;Saltikoff et al, 2019;Smith et al, 2012). However, climatological characterisations of rainfall patterns during HPEs are rare in literature and often based on rain gauge identification of HPEs (Panziera et al, 2018;Thorndahl et al, 2014).…”

A multi‐year assessment of sub‐hourly gridded precipitation for Switzerland based on a blended radar—Rain‐gauge dataset

“…Keupp et al (2017) give a comprehensive overview of existing radar-derived rainfall climatologies. For example, radar-derived rainfall climatologies were produced for the Netherlands (Overeem et al, 2009), for Great Britain and Ireland (Fairman et al, 2015), for France (Tabary et al, 2012), for Germany (Paulat et al, 2008;Brendel et al, 2015), for the Czech Republic (Bližňák et al, 2018), for Denmark (Thorndahl et al, 2014), for Belgium (Goudenhoofdt and Delobbe, 2016), for Sweden (Berg et al, 2016) and for Switzerland (Wüest et al, 2010;Sideris et al, 2014a;Panziera et al, 2018).…”

“…Rain gauges are the best available data to adjust radar data, since they provide accurate point measurements of rainfall and can be considered as the ground truth (Schmidli et al, 2001) despite the fact that also rain gauges can be subject to measurement and representativeness errors (e.g., Groisman and Legates, 1994;Yang et al, 1999;Habib et al, 2001;Schmidli et al, 2001;Molini et al, 2005;Sevruk, 2006;Tapiador et al, 2012;Berndt et al, 2014). Various techniques are used to adjust the raw radar data using ground observations, they include the correction of systematic biases using gauge observations (Overeem et al, 2009;Thorndahl et al, 2014;Brendel et al, 2015;Fairman et al, 2015;Bližňák et al, 2018), incorporating radar rainfall structures through a temporal downscaling step of rain-gauge analysis (Paulat et al, 2008;Wüest et al, 2010) or geostatistical merging (Tabary et al, 2012;Sideris et al, 2014a;Goudenhoofdt and Delobbe, 2016). Evaluations of radar-gauge combination methods show that in general geostatistical merging techniques out-perform mean field bias corrections (Goudenhoofdt and Delobbe, 2009).…”

“…Evaluations of radar-gauge combination methods show that in general geostatistical merging techniques out-perform mean field bias corrections (Goudenhoofdt and Delobbe, 2009). Since discrepancies between radar and rain-gauge data become more important for smaller accumulation times, it is common practice to employ rainfall aggregations over at least 1 hr (e.g., Zawadzki, 1975;Villarini et al, 2008), which are subsequently disaggregated to derive rainfall depths at finer temporal resolutions such as the hourly (e.g., Paulat et al, 2008;Wüest et al, 2010;Tabary et al, 2012;Berg et al, 2016), 30-min (e.g., Bližňák et al, 2018) and 5-min (e.g., Sideris et al, 2014a) time scales.…”