Ocena przydatności disdrometru laserowego i radaru meteorologicznego do szacowania wielkości opadów deszczu.

Contemporary challenges in the management of stormwater and modelling of rainfall--runoff processes (in urban areas in particular) require the use of rainfall-estimation devices more advanced than rain gauges. One such device is the laser disdrometer, which allows (alongside radar reflectivity) for measurement of the intensity of rainfall of high temporal resolution and an accuracy greater than that available using rain gauges. On the other hand, meteorological radar makes it possible to estimate rainfall with a high degree of spatial resolution. The disadvantage of radar observations is the inaccuracy of the rainfall data obtained. Measurements of atmospheric precipitation conducted at the WULS-SGGW Meteorological Station in Warsaw in the years 2012-2014 and 2019-2020, using a tipping-bucket rain gauge and the laser disdrometer (Parsivel), were combined with data obtained from the meteorological radar in Legionowo (in the C-band), with this allowing data to be collected to allow for assessment of the usefulness of the disdrometer and radar where the estimation of rainfall is concerned. The two instruments have independent systems by which to record precipitation data, ensuring a mutual time shift. This made temporary synchronisation a necessity. The data for the entire study period were used in analysing correlations between 24-hour rainfall depths estimated on the basis of the rain gauge, on the one hand; and the disdrometer on the other. The correlation coefficient R obtained was equal to 0.87. However, the total amount of rainfall calculated on the basis of the data from the disdrometer was about 40% greater than the corresponding value from the rain gauge. From the dataset for the years 2012-2014, 21 individual events were selected for further analysis, with these being ones for which radar-estimated rainfall data in the form of a PAC hydrological product generated by the system belonging to Poland's Institute of Meteorology and Water Management (IMGW-PIB) were also available. The data measured using the rain gauge and the disdrometer were characterised by a high time resolution, of 1 min. The rainfall-intensity values obtained from the PAC product had a temporal resolution of 10 min and a spatial resolution of 1 km. The rainfall data from the disdrometer and radar were then used in analysing the correlations between these and corresponding measurements made by rain-gauge. The mean and median values of the R correlation coefficient, obtained in these analyses on the basis of rainfall-intensity values averaged over 10-min time intervals (though observed at the basic 1-min resolution) were, when estimated using the disdrometer, of 0.98 and 0.99 respectively. Correlations based around rainfall-intensity values at the 1-min level of resolution only assumed lower values. The adequate values of the R coefficient, as determined for the radar data, were of 0.68 and 0.77 respectively. The study also extended to include comparison of total values estimated for 21 individual rainfall events using the disdrometer and radar (the PAC hydrological product), as compared with data measured using the rain gauge. The analysed values from the disdrometer were greater than the corresponding rain-gauge values for almost all events, while those obtained using radar were lower in most cases. The mean and median values of the relative error, obtained in relation to the values of rainfall totals measured using the disdrometer, were 35.2 and 38.3% respectively. The relative error values, obtained in adequate analysis based around data from the PAC radar product, proved to be much higher, and amounted to 49.1 and 59.1% respectively. This analysis therefore made it clear how disdrometer- and radar-based data require prior correction before any potential use can be made of them, e.g. in hydrological analyses. This paper's simple method of adjusting the heights of rainfall estimated on the basis of the disdrometer at specific (assumed 10-min) time intervals during the event was able to achieve a significant reduction of differences in the total rainfall values for single events, as supplied by data from the disdrometer and the rain gauge. In regard to the adjusted data from the disdrometer, the mean and median values for relative error were of 15.5 and 17.6% respectively, in respect of the 12 rainfall events used to verify the method. [ABSTRACT FROM AUTHOR]