Please use this identifier to cite or link to this item:
Title: 應用非點源汙染模式SWAT模擬翡翠水庫上游集水區流量及氮素之輸出與移動
Application of nonpoint source pollution model SWAT to evaluate streamflow and the flow pathways of nitrogenous fertilizer in the upstream watershed of Feitsui Reservoir.
Authors: 李宗祐
Lee, Tsung -Yu
Guo, Nae-Wen
Lin, Guan-Zhou
Keywords: 非點源汙染
non-point pollution
SWAT model
nitrate modelling
nitrogen fertilizer
flow pathway
nitrogen pathway
Issue Date: 2017
Abstract: 人為活動發展產生的非點源汙染難以了解其移動途徑而使得控制不易,其中農業活動所使用的肥料為非點源汙染的主要來源之一,翡翠上游集水區茶葉種植對水質的影響,一直是關注的焦點。本研究透過SWAT模式模擬翡翠水庫上游集水區2012年至2014年間日流量與日硝酸鹽氮之輸出量,並藉由SWAT-CUP檢定該集水區肥料施放量,進而從模式模擬之結果了解氮肥於農地移動路徑與分布,藉以評估肥料被茶樹利用之情形。SWAT模式於模擬日流量與河川日硝酸鹽氮輸出量皆達到良好的水準以上(Nash-Sutcliffe model efficiency coefficient>0.65),各集水區肥料檢定成果落於200~420kg/ha,符合農糧署建議之施放量。 本研究先探討水質採樣頻率對於硝酸鹽氮觀測值推估之影響,成果顯示九十天採樣頻率所造成的年推估誤差最大(各集水區誤差最大可達1692%);並從氮素移動路徑分析,得出大部分集水區茶樹吸收量皆少於其輸入總量(含乾濕沉降與肥料)之60%,顯示有40%以上輸入的氮素儲存在土壤、入滲至地下水或進入河川;翡翠水庫上游集水區茶園氮素單位面積輸出量高於森林約3~7倍,茶園地表逕流氮素輸出量為森林的5~25倍,茶園側向流氮素輸出量為森林的1~8倍,茶園土壤氮素輸入量為森林的2~8倍,茶園地下水輸出量為森林2~6倍,顯示人為農地施肥對於河川與土壤將有所影響。肥料減半情境之模擬結果顯示,在收穫減少約10~20% (約60~100kg/ha之產量)之情況下,各氮素移動路徑可削減約50%之輸出量。本研究驗證了SWAT模式在臺灣集水區的適用性,未來配合經濟模式的評估,將可綜觀模擬農業策略對於集水區環境與農業產值之影響,做為集水區經營規劃的參考依據。
Human activity-induced nonpoint source pollution cannot be well controlled owing to its unpredictable pathways. Fertilizer is one of the most important nonpoint pollution sources in the cultivated watersheds. Tea cultivation located in the upstream of the reservoir would hence pollute the water of the reservoir. This study used SWAT to simulate daily runoff and nitrate flux in the upstream watersheds of the Feitsui Reservoir in 2012-2014. Besides, SWAT-CUP was used to calibrate fertilizer amount. Through the investigation of modeling results, we wanted to identify the flow pathways of nitrogenous fertilizer and evaluate the efficiency of fertilization by comparing the amount of fertilizer with the amount of nitrogen that uptake by tee trees. SWAT can simulate the observed daily runoff and nitrate flux well (Nash-Sutcliffe model efficiency coefficient >0.65). The result showed that during the period of 2012 to 2014 the amount of applied fertilizer was between 200kg-N/ha to 420kg-N/ha in each watershed. This study investigate the water quality sampling frequency how to influence the nitrate flux estimation of daily simulated .The result show the ninety days frequency will cause the most deviation of the yearly nitrate flux estimation ( most deviation is 1682%) .We analyze the nitrate pathways demonstrate the uptake of tea tree was less than 60% of the total nitrogen input (including dry/wet deposition and fertilizer), indicating more than 40% of the nitrogenous fertilizer was either flushed off to the stream or stored in the watershed, e.g. in soil or to groundwater. The agriculture land exported nitrogen was 3-7 times of the forest. The tea farm surfaceflow exported nitrogen was 5~25 times of the forest and the lateral flow exported nitrogen was one 1~8 times of the forest. The tea farm of nitrogen input to soil was 2~8 times of forest and the tea farm groundwater exported nitrogen was 2~6 times of forest which demonstrate the human activity will influence the soil and river We set the situation of fertilizer application half to simulate the nitrogen pathways and tea harvest. The result demonstrated that the harvest will loss about 10~20% (about 60~100kg/ha), but it could loss about 50% nitrogen of each pathways. We expect to coordinate the nitrogen stored in soil and the tea trees mature to make fertilizer half strategy and analyze the influence of economy and watershed environment. We hope that this strategy can promote to the other watershed of eutrophication in the future.
Other Identifiers: G060323022L
Appears in Collections:學位論文

Files in This Item:
File SizeFormat 
060323022l01.pdf10.81 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.