The Thermodynamic Cloud Phase value-added product (THERMOCLDPHASE VAP) is developed to provide vertically resolved cloud hydrometeor thermodynamic phase identification. Thermodynamic cloud phase identification is important to understand many cloud processes and to retrieve cloud properties. This VAP uses the multisensor approach developed by Shupe (2007) to combine measurements from active remote-sensing instruments including lidars and radars, passive remote-sensing instruments such as microwave radiometers, and radiosondes to determine vertically resolved thermodynamic cloud phases. The VAP reads micropulse or high-spectral-resolution lidar backscatter and depolarization ratio data from the Micropulse Lidar Cloud Mask (MPLCMASK) VAP; radar reflectivity, mean Doppler velocity, and Doppler spectra width data from the Active Remote Sensing of CLouds (ARSCL) VAP; liquid water path data from the Microwave Radiometer Retrievals (MWRRET) VAP; and temperature data from the Interpolated Sonde (INTERPSONDE) VAP. THERMOCLDPHASE then outputs seven hydrometeor phase classifications at 30-meter vertical and 30-seconds temporal resolutions.
Vertically resolved cloud hydrometeor thermodynamic phase identification can be used to improve cloud macrophysical and microphysical property retrievals because most retrieval algorithms are developed for a specific cloud phase and type (Shupe et al. 2016). The data can also be used to study cloud processes such as cloud thermodynamic phase evolution, mixed-phase cloud processes, and cloud life cycle, and to improve our understanding of cloud radiative properties and atmospheric radiative budget (de Boer et al. 2011). Furthermore, the data can be used to validate and improve model simulations of clouds, especially mixed-phase clouds.
The classification thresholds and constraints used in the VAP are determined based on our knowledge of cloud physical properties and existing literature. However, quantitatively assessing the uncertainty of the classification algorithm is challenging because of the absence of a definitive validation dataset. Potential sources of error include the temporal interpolation of soundings, the uncertainty associated with liquid water path retrieval, and the thresholds used in the classification algorithm. These errors particularly affect clouds near temperature thresholds and thin clouds with minimal liquid water. Evaluations of the classification are necessary.