應用衛星影像辨識與河道水理演算於濁水溪沖積扇地下水數值模擬

Applying Satellite Image Identification Technology and River Channel Routing Module to Groundwater Simulation of Chou-Shui River

楊深惠(Shen-Hui Yang)

指導教授 ： 張良正

國立交通大學/工學院/土木工程系所/碩士(2012年)

https://doi.org/10.6842/NCTU.2012.00389

摘要

濁水溪沖積扇為台灣地區最重要的地下水區之一，因此以往已有不少濁水溪沖積扇相關地下水數值模擬之研究，惟多未對濁水溪本流與沖積扇地下水系統間之交換進行較深入之探討。因此，為了提昇對濁水溪本流對區域地下水影響之掌握，本研究以河道模擬演算及衛星影像辨識，分別估計濁水溪之河水位高程與河道水面範圍變化，以作為地下水數值模式中河道模組(RIV package)之關鍵參數，並以RIV package探討濁水溪本流與沖積扇地下水系統間之交換關係。在河道水理演算部分，本研究以2008年之彰雲大橋、西濱大橋、自強大橋及溪州大橋之月平均水位資料與月平均流量資料，及2007年12月之河川量測斷面進行河道水理模式之設定與檢定，並模擬全河道之水位。在影像辨識部分，分別以2008年一月、七月與十一月之福衛二號多光譜之衛星影像圖，並採其中之近紅外光波段作為辨識河道之依據。
研究結果顯示，在河道水理演算方面，濁水溪下游之西濱大橋，其水深於枯水期與豐水期時水深為0.3及1.7公尺，相差1.4公尺；而在濁水溪上游彰雲橋水深較深，其枯水期與豐水期時水深為2.2及0.2公尺，豐枯水期水深相差至2公尺。衛星影像辨識發現，濁水溪於2008年一月、七月與十一月之辮狀河道面積分別為18.95、31.08與17.51平方公里，枯水期之河道面積約僅佔豐水期面積之60%，綜合前述，豐水期有較高之河水位與河道面積且與估水期差距甚大。而地下水數值模式方面，則推估出濁水溪之河川入滲量為每年5.26億噸、河川出滲量為每年1.12億噸、整體沖積扇之補注量為每年20.23億噸，抽水量為每年19.3億噸，比較各種相關研究文獻，顯示此抽水量值應屬合理範圍。

關鍵字

衛星影像辨識 ； 河道模擬 ； 地下水模擬

並列摘要

Many groundwater simulation models have been developed for Chou-Shui River alluvial fan, one of the most important groundwater area. However, the exchange between Chou-Shui River and the groundwater system of the alluvial fan is much less studied. This exchange is evaluated using a river package in the groundwater simulation model. Several critical parameters used in the river package are determined using the relationship between the river water level and the area of river flow surface. This relationship is evaluated by satellite image technology. The river package is calibrated using the monthly average water level data and the monthly average flow data of year 2008 at Chang-Yun Bridge, Xi-Bin Bridge, Chi-Chiang Bridge and Si-Jou Bridge. The field measurement of the cross-section in December 2007 is used in the river package. For calibration of satellite image technology, the FORMOSAT multispectral satellite images of January, July, and November in 2008 are used. The Infrared light is used to identify the river channel.
The results show that the water level differences between wet and dry periods are 1.4m and 2.0m for Xi-Bin Bridge station and Chang-Yun Bridge station respectively. Specifically, 0.3m during dry period and 1.7m during wet period at Xi-Bin Bridge station and 0.2m during dry period and 2.2m during wet period at Chang-Yun Bridge station. The identified areas of the braided channel in Chou-Shui River using the satellite image technology are 18.95, 31.08, and 17.51 km2 in January, July, and November of year 2008 respectively. The area during the dry periods is about 60% of the area during the wet periods. The exchange rates between the river and the groundwater system are 526 million cubic meters and 112 million cubic meters for flowing out and into the river. The total recharge is 20.23 million cubic meters. The pumping rates is 19.3 million cubic meters.

並列關鍵字

Satellite Image Identification ； River Channel Routing Module ； Groundwater Simulation

參考文獻

33. 李振誥、葉信富、謝壎煌，應用河道水位變化評估新虎尾溪地下水補注量之研究，農業工程學報第53卷第2期，2007。

39. 陳韋圻，應用專家系統於地下水模式自動化參數檢定之研究，國立交通大學，碩士論文，民國97年。

40. 王雲直，應用專家系統於地下水模式參數檢定之研究─以濁水溪沖積扇為例，國立交通大學，碩士論文，民國99年。

2. Butler, J. J., Jr., Zlotnik, V.A., and Tsou, M. S., 2001, Drawdown and stream depletion produced by pumping in the vicinity of a partially penetrating stream, Ground Water, vol.39, no.55, pp.651-659.

3. Chau, K.W., 2003. Manipulation of numerical coastal flow and water quality models. Environmental Modelling and Software 18 (2), 99–108.