How Much Can We Learn About Rain Microphysics from Polarimetric Radar Observations? An Investigation of Information Content and Parameter Estimation Using the Bayesian Observationally-constrained Statistical-physical Scheme (BOSS)
Abstract
Bulk microphysics schemes exhibit deficiencies that are due in part to their simplified representation of a complex natural state, and in part due to a fundamental lack of understanding of microphysical processes. Polarimetric radar observations provide insight into the microphysical evolution of clouds, but alone they are unable to provide quantitative information about the process rates. The Bayesian Observationally-constrained Statistical-physical Scheme (BOSS) bridges this gap, allowing information on microphysical processes to be gained by models from observations. BOSS operates without a predefined drop size distribution (DSD) shape and makes no assumptions about parameterized process rates. Instead, BOSS uses a Markov Chain Monte Carlo sampler within a Bayesian inference framework to constrain model microphysics directly with radar observations. A new moment-based polarimetric forward operator is used to relate model prognostic moment output to polarimetric radar variables. The prognostic moment values initialized at model top are retrieved from radar observations utilizing a new probabilistic rain property retrieval. BOSS has the flexibility to choose the type and number of prognostic DSD moments predicted. In this study, we use this flexibility to explore the information content gained by profiles of polarimetric radar variables (ZH, ZDR, KDP) when BOSS predicts different combinations of the 0th, 3rd and 6th DSD moments. Polarimetric profiles are synthetically generated using bin model simulations. When compared to constraining BOSS with moment fluxes alone, we show that constraining with profiles of radar information improves the constraint of process rates, thereby leading to improved rain rate estimation. The polarimetric variable that produces the most information gain varies among simulations and process rates, suggesting that each radar variable provides unique information to BOSS. The inclusion of higher order prognostic moments (e.g. the 6th moment) is also shown to improve process rate constraint and rain rate estimation due to the better correlation between higher order moments and the polarimetric radar variables used to constrain BOSS. We also display initial results of BOSS constrained by real polarimetric radar observations within a limited subset of warm rain cases.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.A43C..06R
- Keywords:
-
- 3310 Clouds and cloud feedbacks;
- ATMOSPHERIC PROCESSES;
- 3311 Clouds and aerosols;
- ATMOSPHERIC PROCESSES;
- 3333 Model calibration;
- ATMOSPHERIC PROCESSES;
- 3354 Precipitation;
- ATMOSPHERIC PROCESSES