An operational, single-polarized X-band weather radar (WRX) provides measurements in Hamburg’s city center since 2013. This local area weather radar (LAWR) is located on the rooftop of the high-rise building "Geomatikum" in Hamburg (HHG), which is the location of the Meteorological Institute of the Universität Hamburg. The radar operates at one elevation angle with a high temporal 30 s, range 60 m, and sampling 1° resolution refining observations of the German nationwide C-band radars within a 20 km scan radius.
Several sources of radar-based errors were adjusted gradually improving the rainfall rate, e.g. the radar calibration, alignment, attenuation, noise, non-meteorologial echoes. This data set group includes monthly data sets of hourly netCDF-files of the rainfall rate (rr) in level 2. The unit is mm/h. Attenuation correction was already applied. The measurement data is available on the original polar grid (360 azimuth angles and 333 range gates) with it's center at the radar location. Latitude and longitude coordinates are included within the files. The variable 'pia_stability' is a quality flag of the stability of the applied attenuation correction affecting the rainfall rate. The level 2 rainfall rate was generated from the radar reflectivity factor (dbz) in level 1 (see data set group "Reflectivities of multi-year X-band weather radar observations in Hamburg (LAWR HHG level 1 dbz)"). The rainfall rate was retrieved from the attenuation-corrected reflectivity using a power-law relationship between these two quantities (Z-R relationship). The standard Marshall-Palmer relationship was applied. The rainfall rate was evaluated to measurements of micro rain radars (MRR). The rainfall rates of WRX/LAWR HHG and MRRs are in good agreement.
These data sets of rainfall rates facilitate refined studies on the spatial and temporal scale of precipitation and further hydrological research, e.g. input data for high-resolution modelling, in an urban area.
Please note that some measurement times are missing due to maintenance and technical errors of the radar. Missing measurement times are not temporal interpolated nor included within the data sets with filled values. The data availability of the years is: 2013 = 57 %, 2014 = 75 %, 2015 = 97 %, 2016 = 98 %, 2017 = 93 %, 2018 = 96 %, 2019 = 98 %, 2020 = 79 %, 2021 = 61 %.
Changes in Version 2:
- We provide daily instead of hourly files to reduce the number of files for better data handling. For the days 23.09.2014, 12.03.2015, 09.06.2015, 05.07.2017, and 01.02.2018 there are two files to avoid additional time dependencies of variables because of changes in calibration or alignment parameters.
- We changed the data type (double to int64) and the unit days since 1970-01-01 to seconds since 1970-01-01 of the time coordinate.
- We changed the standard names / long names of the variables azimuth, range and ele.
- We added the integer variable grid_mapping with the attributes grid_mapping_name ("radar_lidar_radial_scan"), latitude_of_projection_origin, longitude_of_projection_origin and height_of_projection_origin, as suggested by the CfRadial conventions. Since the grid_mapping variable provides the same information as the variables lat_center, lon_center and zsl_center, we removed them. We added the attribute grid_mapping to the variable rr and dbz.