運用SOM和CNN建置以淹水時空變化特性為基礎之區域淹水預報模式

Abstract

21世紀，在世界各國人口持續成長及經濟快速發展的趨勢下，許多國家面臨嚴重的都市化問題，加上近年來氣候變遷日漸加劇，導致極端水文災害頻傳，暴雨事件不僅降雨時空分布更加不均，降雨強度也相較以往增強許多，又各大都市區域大都集中於低窪且易淹水之沖積平原，故致災程度及受災人口相較以往劇烈與增加許多，因此勢必需要藉由了解集水區的淹水時空分布及變化特性，進而將其應用於即時預報，預測未來時刻的區域淹水情況。 台灣因地形變化劇烈，河川坡陡流急、集流時間短，且各大都會區均坐落於沖積平原區域，受到對流雨、鋒面雨及颱風等挾帶高降雨強度之豪雨，種種水文、地質和氣象因素使得洪水事件發生頻繁且難以掌控。近年來，面對人工智慧趨勢崛起，機器學習方法之快速及準確性已被證明能夠有效應用於解決許多即時的氣候災害議題。 本研究使用自組特徵映射網路(SOM)、K-means聚類法及卷積神經網路(CNN)針對臺南市二仁溪流域之45545個40×40m的淹水網格點進行區域淹水變化量之預報。SOM可將高維度的歷史淹水事件以聚類的方式擷取淹水特徵，進而映射至二維特徵拓樸圖上，其拓樸結構有助於探討集水區發生不同程度淹水之時空變化特性；K-means聚類法則用於根據淹水時空變化特徵將集水區進行分區；CNN可藉由卷積運算擷取時間序列資訊之特徵，進而進行未來1~3小時之淹水變化量之預報。 本研究認為使用4×4的SOM網路大小，並於形成次序階段及收斂階段分別迭代2500及500次之結果最能完整涵蓋二仁溪流域之淹水變化特徵。以預測未來一小時變化量(T+1)模式為例，本研究比較全區預測模式及分區預測模式結果，評估指標結果MAE分別為0.038和0.034及RMSE為0.159和0.151，均顯示使用SOM拓樸圖權重進行分區確實能提高預報精準度、降低累積誤差量。SOM-CNN套配結果亦顯示分區模式能有效降低大部份分區之誤差指標。 研究結果顯示結合SOM、K-means及CNN進行區域淹水變化預報方法能快速掌握區域的淹水特徵及淹水變化趨勢，本研究提出的方法可為決策者和當地居民提供區域淹水的時空資訊，進而針對防範洪水採取預防措施。

The frequency of extreme hydrological events caused by climate change has increased in recent years. Besides, most of the urban areas in various countries are located on low-lying and flood-prone alluvial plains such that the severity of flooding disasters and the number of affected people increase significantly. Therefore, it is imperative to explore the spatio-temporal variation characteristics of regional floods and apply them to real-time flood forecasting. Flash floods are common and difficult to control in Taiwan due to several geo-hydro-meteorological factors, including drastic changes in topography, steep rivers, short concentration time, and heavy rain. In recent decades, the emergence of artificial intelligence (AI) and machine learning techniques has proven to be effective in tackling real-time climate-related disasters. This study combines an unsupervised and competitive neural network, the self-organizing map (SOM), K-means clustering, and the convolutional neural networks to make regional flood inundation forecasts. The SOM can be used to cluster high-dimensional historical flooding events and map the events onto a two-dimensional topological feature map. The topological structure displayed in the output space is helpful in exploring the characteristics of the spatio-temporal variation of different flood events in the investigative watershed. The K-means clustering is used to regionalize the river basin using the spatio-temporal characteristic of inundation variation. The convolutional neural networks are suitable for forecasting time-vary systems because its convolution mechanism can extract the feature of time series data, and predict the variation of t+1~t+3. The results demonstrate that the real-time regional flood inundation forecast model combining SOM, K-means, and convolutional neural networks can more quickly extract the characteristics of regional flood inundation and more accurately produce multi-step ahead flood inundation forecasts than the traditional methods. The proposed methodology can provide spatio-temporal information of flood inundation to decision-makers and residents for taking precautionary measures against flooding.