Rain type classification into convective and stratiform is an essential step required to improve quantitative precipitation estimations by remote sensing instruments. Previous studies with Micro Rain Radar (MRR) measurements and subjective rules have been performed to classify rain events. However, automating this process by using machine learning (ML) models provides the advantages of fast and reliable classification with the possibility to classify rain minute by minute. A total of 20,979 min of rain data measured by an MRR at Das in northeast Spain were used to build seven types of ML models for stratiform and convective rain type classification. The proposed classification models use a set of 22 parameters that summarize the reflectivity, the Doppler velocity, and the spectral width (SW) above and below the so-called separation level (SL). This level is defined as the level with the highest increase in Doppler velocity and corresponds with the bright band in stratiform rain. A pre-classification of the rain type for each minute based on the rain microstructure provided by the collocated disdrometer was performed. Our results indicate that complex ML models, particularly tree-based ensembles such as xgboost and random forest which capture the interactions of different features, perform better than simpler models. Applying methods from the field of interpretable ML, we identified reflectivity at the lowest layer and the average spectral width in the layers below SL as the most important features. High reflectivity and low SW values indicate a higher probability of convective rain.

When meteorological conditions deviate from the optimal range for human well-being, the risks of illness, injury, and death increase, and such impacts are feared in particular with more frequent and intense extreme weather conditions resulting from climate change. Thermal indices, such as the universal thermal climate index (UTCI), can better assess human weather-related stresses by integrating multiple weather components. This paper quantifies and compares the seasonal and spatial association of UTCI with mortality, morbidity, and road accidents in the federal state of Bavaria, Germany. Linear regression was applied to seasonally associate daily 56 million hospital admissions and 2.5 million death counts (1995–2015) as well as approximately 930,000 road accidents and 1.7 million people injured (2002–2015) with spatially interpolated same day- and lagged- (up to 14 days) average UTCI values. Additional linear regressions were performed stratifying by age, gender, region, and district. UTCI effects were clear in all three health outcomes studied: Increased UTCI resulted in immediate (1–2 days) rises in morbidity and even more strongly in mortality in summer, and lagged (up to 14 days) decreases in fall, winter, and spring. The strongest UTCI effects were found for road accidents where increasing UTCI led to immediate decreases in daily road accidents in winter but pronounced increases in all other seasons. Differences in UTCI effects were observed e.g. between in warmer north-western regions (Franconia, more districts with heat stress-related mortality, but hospital admissions for lung, heart and external reasons decreasing with summer heat stress), the touristic alpine regions in the south (immediate effect of increasing UTCI on road accidents in summer), and the colder south-eastern regions (increasing hospital admissions for lung, heart and external reasons in winter with UTCI). Districts with high percentages of elderly suffered from higher morbidity and mortality, particularly in winter. The influences of UTCI as well as the spatial and temporal patterns of this influence call for improved infrastructure planning and resource allocation in the health sector.

Climate change and increasing risks of extreme weather events affect human health and lead to changes in the emergency department (ED) admissions and the emergency medical services (EMS) operations. For a better allocation of resources in the healthcare system, it is essential to predict ED numbers based on environmental variables. This publication aims to quantify weather, air pollution and calendar-related effects on daily ED admissions. Analyses were based on 575,725 admissions from the web-based IVENA system recording all patients in the greater Munich area with pre-hospital emergency care in ambulance operations during 2014-2018. Linear models were used to identify statistically significant associations between daily ED admissions and calendar, meteorological and pollution factors, allowing for lag effects of one to three days. Separate analyses were performed for seasons, with additional subset analyses by sex, age and surgical versus internal department. ED admissions were exceptionally high during the three-week Oktoberfest, particularly for males and on the weekends, as well as during the New Year holiday. Admissions significantly increased during the years of study, decreased in spring and summer holidays, and were lower on Sundays while higher on Mondays. In the warmer seasons, admissions were significantly associated with higher temperature, adjusting for the effects of sunshine and humidity in all age groups except for the elderly. Adverse weather conditions in non-summer seasons were either linked to increasing ED admissions (from storms, gust) or decreasing them from rain. Mostly, but not exclusively, in winter, increasing ED admissions were associated with colder minimum temperatures as well as with higher NO and PM10 concentrations. In addition to standard calendar-related factors, incorporating seasonal weather, air pollutant and interactions with patient demographics into resource planning models can improve the daily allocation of resources and staff of EMS operations at hospital and city levels.

Rain properties vary spatially and temporally for several reasons. In particular, rain types (convective and stratiform) affect the rain drop size distribution (DSD). It has also been established that local weather conditions are influenced by large-scale circulations. However, the effect of these circulations on rain microstructures has not been sufficiently addressed. Based on DSD measurements from 16 disdrometers located in Lausanne, Switzerland, we present evidence that rain DSD differs among general weather patterns (GWLs). GWLs were successfully linked to significant variations in the rain microstructure characterized by the most important rain properties: rain intensity (R), mass weighted rain drop diameter (D_m), and rain drop concentration (N), as well as Z = AR^b parameters. Our results highlight the potential to improve radar-based estimations of rain intensity, which is crucial for several hydrological and environmental applications.

Rain microstructure parameters assessed by disdrometers are commonly used to classify rain into convective and stratiform. However, different types of disdrometer result in different values for these parameters. This in turn potentially deteriorates the quality of rain type classifications. Thies disdrometer measurements at two sites in Bavaria in southern Germany were combined with cloud observations to construct a set of clear convective and stratiform intervals. This reference dataset was used to study the performance of classification methods from the literature based on the rain microstructure. We also explored the possibility of improving the performance of these methods by tuning the decision boundary. We further identified highly discriminant rain microstructure parameters and used these parameters in five machine-learning classification models. Our results confirm the potential of achieving high classification performance by applying the concepts of machine learning compared to already available methods. Machine-learning classification methods provide a concrete and flexible procedure that is applicable regardless of the geographical location or the device. The suggested procedure for classifying rain types is recommended prior to studying rain microstructure variability or any attempts at improving radar estimations of rain intensity.