本文選題：數(shù)字流域 + 時空特征��； 參考：《浙江大學(xué)》2014年博士論文

【摘要】：數(shù)字流域特征是對自然界流域特征的數(shù)字化表達(dá),是流域水文模型的數(shù)據(jù)基礎(chǔ),對于模擬流域水文過程、保護(hù)與治理流域生態(tài)環(huán)境具有重要的作用。因此,數(shù)字流域特征提取是地理信息系統(tǒng)領(lǐng)域的一個重要問題。 隨著測繪技術(shù)、計算機技術(shù)、遙感技術(shù)的不斷發(fā)展,基于DEM的數(shù)字流域特征提取取得了較大的進(jìn)展。但由于自然界紛繁復(fù)雜,目前的研究還不能很好地反映自然界的實際流域情況,體現(xiàn)在：無法反映因河流分叉而形成的具有多重歸屬的流域情況；無法反映受時相影響而導(dǎo)致河流流向動態(tài)變化所形成的具有動態(tài)歸屬的流域情況(流域時相特征)；沒有深入研究內(nèi)流流域特征及其提取方法。此外,現(xiàn)有的數(shù)字流域特征提取方法效率較低,尤其是在海量數(shù)據(jù)處理時,效率問題甚至成為了數(shù)字流域分析的瓶頸問題。 本文對數(shù)字流域分析中的洼地填平、流向分析、匯流分析、虛擬水系提取以及流域劃分等問題進(jìn)行了全面的研究,并重點對以下五個方面進(jìn)行了深入的研究： (1)針對由河流分叉所造成的具有多重歸屬屬性的流域,本文將這種流域定義為公共流域,并提出了公共流域的提取方法：首先對河流進(jìn)行拓?fù)潢P(guān)系構(gòu)建,然后對河流歸屬地進(jìn)行地理編碼,并將編碼值賦予直接匯入歸屬地的河流；在此基礎(chǔ)上,以被賦予編碼的河流為“種子河流”,根據(jù)河流之間的拓?fù)潢P(guān)系,將編碼值從下游河流傳遞至上游河流,確定具有多重歸屬屬性的河流；最后結(jié)合從DEM中提取的流向信息,提取公共流域。 (2)針對河流流向動態(tài)變化所造成的具有動態(tài)歸屬屬性的流域,本文將這種流域定義為動態(tài)流域,并提出了動態(tài)流域提取方法：將相互連通的流向動態(tài)變化的河流集合抽象為“節(jié)點”,與流向固定的河流共同構(gòu)成河流網(wǎng)絡(luò)拓?fù)浣Y(jié)構(gòu)；在此基礎(chǔ)上,對河流歸屬地進(jìn)行地理編碼,并將編碼值賦予直接匯入歸屬地的河流；根據(jù)拓?fù)潢P(guān)系,將編碼值從下游河流傳遞至上游河流,確定具有動態(tài)歸屬屬性的河流；最后結(jié)合從DEM中提取的流向信息,提取動態(tài)流域。 (3)針對因地表徑流無法流入海洋而形成的內(nèi)流流域,本文根據(jù)影響內(nèi)流流域的主要因素,將內(nèi)流流域分為：地形型內(nèi)流流域(在DEM中以洼地的形式存在)和氣候型內(nèi)流流域(在DEM中不以某種特殊的形式存在,但通常包含內(nèi)流河),并提出了內(nèi)流流域提取方法：首先利用實測內(nèi)流河對DEM進(jìn)行地形約束,降低內(nèi)流河所在柵格點高程,增大內(nèi)流河與周圍區(qū)域的高程差,使內(nèi)流河所在區(qū)域形成洼地,統(tǒng)一氣候型內(nèi)流流域與地形型內(nèi)流流域在DEM中的表現(xiàn)形式(洼地)；然后對地形約束后的DEM進(jìn)行洼地填平,并識別內(nèi)流流域所在平地,識別規(guī)則為：平地中包含內(nèi)流河或平地中所包含的洼地面積超過設(shè)定閩值；進(jìn)而將內(nèi)流流域所在平地的高程恢復(fù)至洼地填平前(將平地退化為洼地),并以洼地為內(nèi)流流域種子區(qū)域,結(jié)合從DEM中提取的流向信息,提取內(nèi)流流域。 (4)針對數(shù)字流域特征提取中洼地與平地處理效率較低的問題,提出了快速洼地與平地處理方法：該方法在洼地填平的同時,記錄水流從洼地(平地)內(nèi)部以最短路徑流至出口的流向,在平地增高處理中,利用該流向以水流追蹤的方式快速確定平地內(nèi)部柵格點與出口的距離；并且,采用距離轉(zhuǎn)換計算平地內(nèi)部柵格點與邊界的距離；從而大幅減少了隨機搜索與迭代處理,提高了處理效率。 (5)針對數(shù)字流域特征提取中匯流分析效率較低的問題,提出了快速匯流分析方法：該方法首先搜索得到流域出水口,然后以出水口為種子點,以流向的反向為搜索方向,向流域內(nèi)部進(jìn)行搜索,構(gòu)建流域“匯流樹”(匯流樹的葉節(jié)點為水流源頭,根節(jié)點為出水口)；最后通過對匯流樹進(jìn)行反向遍歷計算匯流累積量。該方法對目標(biāo)區(qū)域內(nèi)的流域進(jìn)行分而治之,縮小了每一個流域匯流分析的搜索范圍,高效利用了內(nèi)存,并且,通過匯流樹有效避免了匯流累積量計算中對數(shù)據(jù)的隨機搜索,從而提高了匯流分析的效率。 本文的主要創(chuàng)新點包括： (1)首次提出了公共流域及其提取方法。 (2)首次提出了動態(tài)流域及其提取方法。 (3)首次對基于DEM的內(nèi)流流域提取方法展開深入研究。 (4)提出了快速洼地與平地處理方法、快速匯流分析方法,提高了數(shù)字流域特征提取的效率。 實驗結(jié)果表明,本文所提出的方法能夠有效提取公共流域、動態(tài)流域、內(nèi)流流域及樹狀流域特征,并且提高了數(shù)字流域特征提取的效率。
[Abstract]:The characteristic of digital basin is the digital expression of the characteristics of the natural basin. It is the data base of the hydrological model of the basin. It plays an important role in simulating the hydrological process of the basin and protecting and controlling the ecological environment of the basin. Therefore, the feature extraction of digital watershed is an important problem in the field of geographic information system.
With the continuous development of Surveying and mapping technology, computer technology and remote sensing technology, the feature extraction of Digital Watershed Based on DEM has made great progress. However, because of the complex nature of nature, the current research can not well reflect the actual situation of the natural river basin, which can not reflect the multiple attribution caused by the branching of the river. River basin conditions; the dynamic attribution of river basin conditions (the characteristics of basin time facies) that can not reflect the dynamic changes in the flow direction of the river, and no in-depth study of the characteristics and extraction methods of the internal flow basin. In addition, the existing method of extracting the characteristics of the digital watershed characteristics is low, especially in the process of mass data processing. The problem has even become a bottleneck problem in digital watershed analysis.
This paper makes a comprehensive study on the problems of low-lying land filling, flow analysis, confluence analysis, virtual water system extraction and watershed division in digital basin analysis, and focuses on the following five aspects.
(1) aiming at a river basin with multiple attribution attributes, this paper defines the basin as a public basin, and puts forward the method of extracting the public basin: first, the topology of the river is constructed, then the geographical coding of the River belongs to the river, and the coding value is given to the river which is directly remitted to the home area; On the basis, the coded rivers are used as "seed rivers". According to the topological relations between rivers, the coded values are transferred from the downstream rivers to the upstream rivers, and the rivers with multiple attribution attributes are determined. Finally, a public basin is extracted from the flow information extracted from the DEM.
(2) in view of the dynamic attribution of river basins, this paper defines the watershed as a dynamic basin, and proposes a dynamic watershed extraction method, which abstracts the set of rivers which are connected to dynamic changes as "nodes", and forms a river network topology together with rivers with fixed flow direction. On this basis, geo coding is carried out on the site of the river, and the coded values are assigned to the river that directly remittance to the land. According to the topological relation, the coded values are transferred from the downstream rivers to the upstream rivers to determine the dynamic attribution of rivers. Finally, the dynamic Basin is extracted from the flow information extracted from the DEM.
(3) based on the main factors that affect the inflow of surface runoff into the ocean, this paper divides the inner stream basin into the main factors that affect the internal flow basin, including the topographic basin (in the form of a depression in DEM) and the climate type basin (not in a particular form in the DEM, but usually included in the inland river). The method of extracting the inner stream basin: first, using the measured inland river to restrain the DEM, reduce the elevation of the grid point of the inland river and increase the elevation difference between the inland river and the surrounding area, make the region of the inland river form Wa Di, unify the manifestation of the climate type internal flow basin and the topographic basin in the DEM, and then to the terrain. The restricted DEM is filled with the depression and identifies the flat land in which the internal flow basin is located. The recognition rule is that the area contained in the inland river or the flat land is more than the setting of the min value, and then the Gao Cheng in the flat land of the inner stream basin is restored to the bottomland (depressions will be degraded to the low-lying land), and the depression is the seed area of the inner flow basin. The flow basin is extracted from the flow information extracted from DEM.
(4) in view of the low efficiency of low-lying land and flat land processing in the feature extraction of digital basin, a method of fast depressions and flat ground treatment is proposed. This method is used to record the flow of water from the shortest path to the outlet flow from the low-lying land (flat ground), while the flow is traced fast in the way of water flow tracing. At the same time, the distance between the grid point and the exit of the interior is determined, and the distance between the grid points and the boundary is calculated by the distance conversion, thus the random search and iterative processing are greatly reduced, and the processing efficiency is improved.
(5) in order to solve the problem of low converge analysis efficiency in the feature extraction of digital watershed, a fast converge analysis method is proposed: this method first searches the water outlet of the basin, then takes the outlet as the seed point, and then searches the river basin with the reverse direction of the flow direction, and constructs a watershed "confluence tree" (the leaf node of the confluence tree is water). Flow source, root node is a water outlet); finally, the sink is calculated by reverse traversal of the confluence tree. This method divides the basin in the target area, reduces the search scope of the flow analysis in each basin, efficiently uses the memory, and avoids the data in the confluence accumulation calculation effectively through the confluence tree. The random search improves the efficiency of the confluence analysis.
The main innovation points of this article include:
(1) the public basin and its extraction methods have been put forward for the first time.
(2) the dynamic basin and its extraction methods are first proposed.
(3) for the first time, an in-depth study on the extraction method of DEM based inflow watershed was conducted.
(4) put forward the method of rapid depressions and leveling, fast convergence analysis method, and improve the efficiency of digital watershed feature extraction.
The experimental results show that the proposed method can effectively extract the characteristics of public basin, dynamic basin, internal flow basin and tree basin, and improve the efficiency of feature extraction of digital watershed.
【學(xué)位授予單位】：浙江大學(xué)
【學(xué)位級別】：博士
【學(xué)位授予年份】：2014
【分類號】：P208;P333

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 楊錦玲;;基于分形的數(shù)字水系集水面積閾值確定研究[J];測繪科學(xué);2011年04期

2 錢柯健;朱紅春;李發(fā)源;;一種基于DEM匯水累積量的徑流節(jié)點提取方法[J];測繪科學(xué);2012年01期

3 左偉,曹學(xué)章,李碩;基于數(shù)字地形分析的小流域分割技術(shù)[J];測繪通報;2003年05期

4 朱慶,趙杰,鐘正,眭海剛;基于規(guī)則格網(wǎng)DEM的地形特征提取算法[J];測繪學(xué)報;2004年01期

5 劉光,李樹德,張亮;基于DEM的溝谷系統(tǒng)提取算法綜述[J];地理與地理信息科學(xué);2003年05期

6 王建平,任立良,吳益;一種新的DEM填洼處理算法[J];地球信息科學(xué);2005年03期

7 于淼;任立良;;基于DEM模型的新填洼算法[J];地球信息科學(xué)學(xué)報;2009年01期

8 李碩,曾志遠(yuǎn),張運生;數(shù)字地形分析技術(shù)在分布式水文建模中的應(yīng)用[J];地球科學(xué)進(jìn)展;2002年05期

9 李麗,郝振純;基于DEM的流域特征提取綜述[J];地球科學(xué)進(jìn)展;2003年02期

10 秦承志;朱阿興;李寶林;裴韜;周成虎;;基于柵格DEM的多流向算法述評[J];地學(xué)前緣;2006年03期

，

本文編號：1975770