颱洪期間河川流量觀測與推估作業之調適改善研究

流量為河川治理及水資源管理等重要背景資料。臺灣常用現場觀測流量產製水位流量率定曲線並應用於推估流量。然而大部分觀測流量範圍小且值低，導致外延使用的可信度低；其主因為現行觀測技術與作業流程於颱洪期間不易執行現場流量觀測作業。為突破觀測困境，水利署設置雷達波表面流速儀，以期可全時觀測與推估流量；但並未建立現場平均與表面流速關係率定、表面流速資料品管與流量推估流程。因此本研究以頭前溪支流油羅溪內灣站為例，採用聲波都卜勒流速剖面儀與定點長時作業方式，提升颱洪期間觀測能力與資料累積效率。同時發展含平均與表面流速關係及表面流速資料品管的指標表面流速法推估颱洪期間完整流量歷線。研究中比較2021年單年8場次與2016至2020間139場次流量觀測，採用聲波都卜勒流速剖面儀與定點長時作業方式使觀測資料數量成長9.39倍，接觸式設備的最大觀測流量則提升18.17倍；亦率定全斷面平均與表面流速比值為0.64，其值與現行地面水文觀測手冊建議值0.86差異大。應用指標表面流速法可有效彌補觀測空缺，並推估2021年間16場颱洪事件流量歷線而達全洪程觀測的效果，其中透過表面流速資料品管作業改善表面流速雜訊達94%。此外研究中亦整合觀測與推估流量產製水位流量率定曲線，如此可擴大產製的資料範圍，並降低水位流量率定曲線外延時因無資料所產生的推估差異。

關鍵字

指標表面流速法；聲波都卜勒流速剖面儀；全洪程觀測；水位流量率定曲線

並列摘要

Discharge is an essential data for river training and water resource managements. In Taiwan, head-discharge rating curve built based on measured discharges is commonly used to estimate discharge. However, high measured discharge data is usually few that leads head-discharge rating curve with low reliability. The two main reasons are improper measurement instruments and inexecutable official standard operation processes applied during flood. Several microwave surface velocity radars were therefore installed by the Water Resource Agency to monitor flow velocity and estimate discharges all time. However, relationship between mean velocity and surface velocity, data quality treatments for measured surface velocity and discharge estimation processes were not considered. In this study, the site Neiwan in the Tochan River watershed was selected as a demonstration site. An acoustic Doppler current profiler was used to measure discharge during flood and the way of long-period measurements at a fixed site was also considered to enhance the measurement ability during flood and accumulation efficiency of measured discharge data, respectively. The Index Surface Velocity Method including relationship between mean and surface velocity and surface velocity data quality treatments was developed in this study to reasonable estimate complete discharge history during flood. In this study, the amount of measured discharge data in 2021 were 9.39 times compared with that from 2016 to 2020 by applying modified discharge measurement operations. The maximum measured discharge grew by 18.17 times in 2021 and the relationship of mean-surface velocity ratio is 0.64 which was highly different from the ratio value 0.86 given from current operation regulations. The Index Surface Velocity Method could efficiently cover the time periods in which discharge could not be measured manually and then approached full-time discharge measurement during flood. In 2021, the discharge histories of 16 flood events were estimated. The developing data quality treatments for measured surface velocity in index surface velocity method could decrease data noises to 94%. In addition, in this study, head-discharge rating curve of site Neiwan was produced based on combination of measured and estimated discharge. The combination could expand the used data range for head-discharge rating curve and decrease deviations of estimated discharges applying extrapolation of head-discharge rating curve produced by measured discharge data only.