Assimilation Experiments of MTSAT Rapid Scan Atmospheric Motion Vectors on a Heavy Rainfall Event

Abstract

Atmospheric motion vectors (AMVs) derived from 5-min rapid scan (RS) imagery of the Multi-functional Transport Satellite are expected to capture small-scale distributions of airflows better than typical AMVs derived from 30-min imagery because the observation interval of RS-AMV is shorter. The impact of these high-frequency data on the numerical forecasting of a heavy rainfall near a stationary front was investigated by conducting data assimilation experiments. As a part of preparation for the assimilation, RS-AMVs were compared with the firstguess field obtained from the Japan Meteorological Agency (JMA) nonhydrostatic model (NHM). The comparison result indicated that the RS-AMVs were of good quality and could be used in the JMA's operational NHM with 4D variational data assimilation (JNoVA). Assimilation experiments investigating a heavy rainfall event were conducted using different lengths of assimilation time slot and time intervals of spatial thinning for the assimilation of the RS-AMV data. The assimilation of RS-AMVs caused the initial wind fields to enhance the upper-level divergence and low-level convergence around the front. Consequently, the forecast of the rainfall amount was increased near the front, and the verification scores were slightly improved over the control experiment in the early forecast hours.

Measurement Performance Assessment of Future Space-Borne Doppler Wind Lidar for Numerical Weather Prediction

“…Bias and root-mean-square (RMS) vector difference of the AMV retrieval verified against radiosonde wind observations as shown in previous studies (e.g. Hayashi and Shimoji 2013;Otsuka et al 2015) are relatively large (bias less than 2 m s −1 and RMS vector difference worse than 4 m s −1…”

Feasibility Study for Future Space-Borne Coherent Doppler Wind Lidar, Part 1: Instrumental Overview for Global Wind Profile Observation

“…As discussed in our previous study (Ishii et al 2016), the AMV achieves a large coverage area and high temporal and horizontal resolutions (2.5 and 10 min, 0.5° × 0.5°) but has a low vertical resolution (2 -4 km). The bias and root-meansquare (RMS) vector difference of the AMV retrieval verified against radiosonde wind observations presented in previous studies (e.g., Hayashi and Shimoji 2013;Otsuka et al 2015) are relatively large (bias of less than 2 m s −1 and RMS vector difference of worse than 4 m s −1 ). The AMV can hardly retrieve vector winds under thick clouds, over dry regions, clear-sky regions or regions with few clouds, the atmosphere near Earth's surface over inland area, and low-windspeed regions.…”

Convective cloud top vertical velocity estimated from geostationary satellite rapid‐scan measurements

“…There have been a number of wide-ranging studies on clouds and precipitation that have taken advantage of geostationary satellite measurements, e.g., investigations of the derivation of the atmospheric motion vector [e.g., Otsuka et al, 2015], detection and monitoring of rapidly developing cumulus areas [e.g., Okabe et al, 2011;Sieglaff et al, 2011]. Furthermore, there have been many studies on the life cycles of cloud-precipitation systems using geostationary satellite measurements [e.g., Chen and Houze, 1997;Rickenbach, 1999].…”