کد مقاله کد نشریه سال انتشار مقاله انگلیسی ترجمه فارسی نسخه تمام متن
4449543 1620498 2017 16 صفحه PDF سفارش دهید دانلود رایگان
عنوان انگلیسی مقاله ISI
Improving precipitation forecast with hybrid 3DVar and time-lagged ensembles in a heavy rainfall event
ترجمه فارسی عنوان
بهبود پیش بینی بارش با 3DVar هیبریدی و گروه تاخیر زمان در رویداد بارش باران های سنگین
دانلود مقاله + سفارش ترجمه
دانلود مقاله ISI انگلیسی
رایگان برای ایرانیان
سفارش ترجمه تخصصی
با تضمین قیمت و کیفیت
کلمات کلیدی
جذب داده؛ ترکیبی؛ آنسامبل زمان تاخیر ؛ پیش بینی بارش
موضوعات مرتبط
مهندسی و علوم پایه علوم زمین و سیارات علم هواشناسی
چکیده انگلیسی


• Performance of hybrid assimilation using time-lagged ensemble is evaluated.
• Flow-dependent covariance is introduced without significantly increasing computation cost.
• Precipitation forecast skill is improved by the proposed method.

This study evaluates the performance of three-dimensional variational (3DVar) and a hybrid data assimilation system using time-lagged ensembles in a heavy rainfall event. The time-lagged ensembles are constructed by sampling from a moving time window of 3 h along a model trajectory, which is economical and easy to implement. The proposed hybrid data assimilation system introduces flow-dependent error covariance derived from time-lagged ensemble into variational cost function without significantly increasing computational cost. Single observation tests are performed to document characteristic of the hybrid system. The sensitivity of precipitation forecasts to ensemble covariance weight and localization scale is investigated. Additionally, the TLEn-Var is evaluated and compared to the ETKF(ensemble transformed Kalman filter)-based hybrid assimilation within a continuously cycling framework, through which new hybrid analyses are produced every 3 h over 10 days. The 24 h accumulated precipitation, moisture, wind are analyzed between 3DVar and the hybrid assimilation using time-lagged ensembles.Results show that model states and precipitation forecast skill are improved by the hybrid assimilation using time-lagged ensembles compared with 3DVar. Simulation of the precipitable water and structure of the wind are also improved. Cyclonic wind increments are generated near the rainfall center, leading to an improved precipitation forecast. This study indicates that the hybrid data assimilation using time-lagged ensembles seems like a viable alternative or supplement in the complex models for some weather service agencies that have limited computing resources to conduct large size of ensembles.

ناشر
Database: Elsevier - ScienceDirect (ساینس دایرکت)
Journal: Atmospheric Research - Volume 183, 1 January 2017, Pages 1–16
نویسندگان
, , , , , ,
دانلود مقاله + سفارش ترجمه
دانلود مقاله ISI انگلیسی
رایگان برای ایرانیان
سفارش ترجمه تخصصی
با تضمین قیمت و کیفیت