Nghiên cứu đồng hóa số liệu địa phương vào mô hình WRF để nâng cao chất lượng dự báo mưa cho khu vực Nam Bộ

0 Người đánh giá. Xếp hạng trung bình 0

Tác giả: Mạnh Dũng Lê, Ngọc Quyền Lê, Kỳ Phùng Nguyễn

Ngôn ngữ: vie

Ký hiệu phân loại:

Thông tin xuất bản: Tạp chí Khí tượng Thủy văn, 2022

Mô tả vật lý: tr.36-51

Bộ sưu tập: Metadata

ID: 329985

Thêm vào giỏ Liên kết toàn văn

In this paper, the author has assimilated local data by using the 3D–var method for the WRF model with a resolution of 3km to forecast rainfall for the Southern VietNam. The data used for assimilation included surface observations, sounding data in Vietnam as well as in Southeast Asia. The experiment was conducted with 2 options with data assimilation (Wrf_d03) and without data assimilation (Wrf_noDA_d03) at lead–time of 48 hours. The implementation time is in July and August 2021 in the rainy season. The forecast results of two experiments were collected with 24 Synop observation stations to evaluate the accuracy of the models. The results show that all models have the rainfall forecast almost higher than actual for light rainfall thresholds, and lower actual for other rainfall thresholds. The MAE and RMSE indexes of Wrf_d03 are mostly lower than Wrf_noDA_d03, this shows that assimilation local data has reduced the error of the model. Wrf_d03 also has POD index, ETS index for moderate and heavy rainfall thresholds better than Wrf_noDA_d03 at the 12h and 24h of lead–times. When running model with data assimilation under Cycling mode will give more stable results than the normal case.Trong nghiên cứu đã sử dụng phương pháp đồng hóa số liệu biến phân ba chiều 3D–var cho mô hình WRF với độ phân giải 3km để dự báo mưa cho khu vực Nam Bộ. Số liệu được sử dụng cho đồng hóa bao gồm các quan trắc bề mặt, thám không ở Việt Nam cũng như khu vực Đông Nam Á. Tác giả tiến hành thử nghiệm với 2 trường hợp có đồng hóa (Wrf_d03) và không có đồng hóa số liệu (Wrf_noDA_d03) với hạn dự báo 48h. Thời gian thực hiện tiến hành trong 2 tháng mùa mưa 7 và 8 năm 2021. Các kết quả dự báo của 2 thử nghiệm được thu thập cùng với số liệu quan trắc 24 trạm Synop trên khu vực để tiến hành đánh giá độ chính xác của các mô hình. Kết quả cho thấy, ở ngưỡng mưa nhỏ các mô hình dự báo thiên cao hơn thực tế. Trong khi đó ở ngưỡng mưa vừa, mưa to đến rất to các mô hình đều có xu hướng dự báo thiên thấp hơn thực tế. Các chỉ số MAE và RMSE của mô hình đồng hóa hầu hết đều thấp hơn mô hình không đồng hóa, cho thấy khi đồng hóa dữ liệu địa phương đã giảm được sai số của mô hình. Với đánh giá dự báo cho ngày 15/7/2021, mô hình có đồng hóa số liệu có khả năng phát hiện (POD) và kỹ năng dự báo (ETS) cho mưa vừa, mưa to tốt hơn mô hình không đồng hóa ở hạn dự báo 12h và 24h. Đồng thời khi thiết lập mô hình đồng hóa chạy ở chế độ Cycling sẽ cho dự báo ổn định hơn trường hợp không đồng hóa..

1. Đọc trực tuyến

1. Đọc trực tuyến

Tạo bộ sưu tập với mã QR