Facilities and Infrastructure
Office & Computing
Our 400 m2 office and indoor laboratory spaces are located in the New York State Center of Excellence in Wireless and Information Technology (CEWIT) building at Stony Brook University. The office and indoor laboratory spaces feature state of the art computer and conference capabilities and access to a rooftop laboratory space for instrument testing. Computing facilities maintained by our research group include a total online server disc capacity of 350 TB. Our three main servers have a total processing power of 96 dual core processors, 5 GPU units and 2 TB RAM.
Radar Facilities
KASPR
The flagship radar of the observatory is a 35-GHz (Ka-band, 8-mm) Scanning Polarimetric Radar (KASPR). KASPR, a state-of-the-art cloud scanning radar, can collect Doppler spectra and radar moments through alternate transmission of horizontally and vertically polarized waves and simultaneous reception of co-polar and cross-polar components of the backscattered wave with the beamwidth of 0.32°, hence a full set of polarimetric radar observables is available. These polarimetric observables allow us to identify microphysical processes.
SKYLER-I
SKYLER-II
ROARS
Software
MAAS
The Multisensor Agile Adaptive Sampling (MAAS, Kollias et al., 2020) cyberinfrastructure (CI), currently supported by the National Science Foundation Directorate for Computer and Information Science and Engineering (Kollias et al., 2020). The MAAS CI’s goal is to significantly improve the ability to sample rapidly evolving atmospheric phenomena by providing better control systems across multiple advanced radar systems (brown box, Figure 4). By better enabling the real-time, fine-grained, coordinated control of atmospheric observing instruments, the MAAS framework can revolutionize the study of convective storms towards the goal of improving our scientific understanding and the prediction of extreme or high-impact storms using physics-based and AI-based models. Transformative elements of the MAAS CI are the use of real-time observations external to the outdoor laboratory (blue box, Figure 4) for improved situational awareness and feature detection and tracking and its potential for integrating future observing technologies, such as drones and phased-array radars (Lamer et al., 2023).
CR-SIM
CR-SIM (the Cloud-resolving model Radar SIMulator) is a forward-modeling framework that converts numerical weather prediction model output into synthetic radar observations by simulating electromagnetic scattering from model-resolved hydrometeors across multiple radar frequencies, polarizations, and viewing geometries (Oue et al., 2020). The simulator accounts for hydrometeor size distributions, phase-dependent dielectric properties, attenuation, and Doppler effects, enabling systematic evaluation of radar observables and sensing strategies; however, in its current form CR-SIM does not explicitly represent partially melted hydrometeors, instead treating ice and liquid species separately, which limits its ability to reproduce realistic melting-layer (bright band) radar signatures. CR-SIM is nevertheless widely used across the meteorological research community due to its open-source availability, modular design, and comprehensive documentation, which have facilitated broad adoption, reproducibility, and extension by both radar and modeling groups.
