| 摘要: | 本研究利用Landsat衛星影像探討近30年(1986-2015)來河內市地表/地物變化對都市熱島強度(UHI)的影響。主要研究目標包括(i)近30年來(1986-2015)河內市都市發展的地表/地物變化，(ii)近30年來地表/地物變化對歸一化植被指數(NDVI)、歸一化差異建築指數(NDBI)、地表溫度(LST)及都市地區降雨之定量分析。河內市為越南之首都，其為都市化成長最快的大都會之一，並位於紅河三角洲中心的西北部，其緯度介於北緯20∘53’至21∘23’，經度介於西經105∘44’至106∘02’。
本研究使用1986、1998及2015年的Landsat衛星影像進行地表/地物分類，而後用ArcGIS隨機選取100個點並轉成KML(Keyhole Markup Language)檔以利在Google Earth中開啟。在本研究中Google Earth當作真實值進行地表/地物分類的準確性評估。除此之外，本研究亦搜集近30年來河內市地面測站的溫度及降雨資料來驗證LST。
研究結果顯示地表/地物分類有相當大的變化，而地表/地物分類在1986、1998及2015年的準確度分別為83%、82%及84%，kappa係數為0.79、0.78及0.8。最後，UHI由LST計算而來，並和地表/地物分類比較。近29年來建物及農業用地面積快速地增加，分別增加282.16平方公里及440.13平方公里。最高的LST由農業用地、建築用地及裸土決定，最低的LST由水體、森林及沙洲類別決定。河內市的LST在近30年間持續增加4.64∘C，同時建物及農業用地也在1986至2015年間明顯的增長，該分類的LST變化分別為4.94∘C及4.74∘C。值得注意的是河內市的NDBI及LST為正相關，而NDVI與LST為負相關。LST藉由四個地面測站的觀測資料進行驗證，有高度的相關性，相關性在1986年、1998年及2015年分別為R^2=0.951、R^2=0.935及R^2=0.939。1948年及2015年的UHI為5.71∘C及7.49∘C。由地面測站降雨資料，都市區域的UHII及降雨為正相關，可推測都市地區降雨增加的趨勢可能是來自於都市熱島效應。人工鋪面(ISA)在都市地區的溫度比郊區高，且兩者於UHII為正相關。
;The impact of land use/ land cover (LULC) changes on urban heat island intensity (UHII), principally caused from urbanization, is investigated with Landsat imagery at Hanoi megacity during the past three decades (1986-2015). For the objective, the primary examinations include (i) the LULC changes caused by urban development in Hanoi during the past three decades (1986 – 2015); (ii) the quantitative analysis of the LULC change impact on Normalized Difference Vegetation Index (NDVI), The Normalized Difference Built-Up Index (NDBI), Land Surface Temperature (LST) and regional precipitation focus on urban growth during past three decades. Hanoi is capital and one of the growing metropolis with the most rapid urbanization speed in Vietnam. It is located in the Northwest of the center of the Red river Delta with coordinates Latitude from 20o 53’ N to 21o 23’ N and Longitude 105o 44’ W and 106o 02’ W.
For this purpose, multi-temporal Landsat images for 1986, 1998, and 2015 were acquired. After I built up LULC classification in 1986, 1998 and 2015, generating a set of random 100 points was done in ArcGIS Map and converting random points to Keyhole Markup Language (KML) in order to open in Google Earth. Each random point’s value verified from Google Earth for accuracy assessment. Google Earth model was used to measure of how many ground truth pixels are correctly classified to validate LULC results. Beside that data ground station about temperature and rainfall also are collected in Hanoi City under period study to validate for LST.
The results of the study showed that LULC in the study area changed significantly with overall classification accuracy of 1986, 1998 and 2015 were 83%, 82% and 84%, with kappa coefficient of 0.79, 0.78 and 0.8, respectively. Finally, UHI effect is determined and the LST values are compared with LULC classes. During 29-year time interval, it was observed that the building area and agricultural land were increased significantly about 282.16 km2 and 440.13 km2, respectively. Maximum LST values were detected for agricultural land, building area, and bare ground, while minimum LST values were detected for water, forest and sandbars classes. The LST is continuously increasing with 4.64 oC in study area over the period under study. Beside that the LST value increases are relatively building and agriculture classes also significantly rose from 1986 to 2015, with the values of 4.94 °C, and 4.74°C, respectively. It was noticed that a positive relationship exists between NDBI and the land surface temperature. On the other side, a negative relationship exists between NDVI and the land surface temperature in Hanoi city. The LST validity of the study results were assessed using real data from four ground stations and the high correlations with R2=0.951, R2=0.935 and R2=0.939 are demonstrated in 1986, 1998 and 2015, respectively. The UHI intensity was computed as 5.71 oC for 1986 and 7.49 oC for 2015. Association the station measurements with UHII, the raising trend in the urban precipitation might be caused by UHI effect, and has positive trend between UHII and fraction rainfall in the urban area. Imperious surface temperature (ISA) value in urban area always higher than rural area and both of them have positive tendency with UHII. |
