Font Size: a A A

Application Research Of “3S” Technology On Grassland Types, Vegetation Coverage And Productivity In Mass Screening

Posted on:2017-03-07Degree:MasterType:Thesis
Country:ChinaCandidate:H Y YuFull Text:PDF
GTID:2283330509451361Subject:Grass science
Abstract/Summary:PDF Full Text Request
Shandan(Containing Shandan Racecourse) is located in the middle of the Hexi Corridor in northern Qilian Mountains, the source of Shiyang river and Heihe river, is the important water conservation area in Hexi Corridor and also the front of north sandstorm. Rational protection and utilization of natural grassland is important to maintain ecological security in Hexi Corridor and sustainable development of intra-regional grassland animal husbandry.In this research, using “3S” combined with ground survey, ≤5m resolution ratio remote sensing data and LANDSAT TM images in the year 1987,1999,2006,2010 and 2014 as data source, using water-temperature index as elevation data, investigated data, ≥0℃ accumulated temperature and wetness degree, evaluating the natural grassland resource, vegetation coverage, productivity and dynamics respectively through decision tree classification, dimidiate pixel model and binary regression model, in order to provide basis and reference for rational protection, construction and utilization of national grassland.The results showed that:(1) Total grassland area in study area were 395900 hm2, alpine meadow, mountain meadow, warm steppe, temperate desert and warm desert grassland were 84700, 31400, 96100, 82300 and 101400 hm2 respectively. Total grassland area decreased 25900 hm2 in 2014 compared with 1987, mountain meadow, warm steppe, temperate desert and warm desert grassland were decreased 8592.75, 2377.08, 3357.69 and 11559.75 hm2 respectively, except no change in the area of alpine meadow, mountain meadow and temperate desert were of significant reduction between 1987 and 1999, temperate desert still showed a trend of decreasing after 1999.(2) Average coverage of each grassland of mountain meadow, alpine meadow, warm steppe, warm desert grassland, warm steppe were decreased one by one, 82.35%, 57.60%, 33.79%, 17.07% and 9.94%, the coverage mainly in class Ⅱ(60%<FVC≤80%),Ⅰ(FVC>80%),Ⅳ(20% < FVC≤40%),V(FVC≤20%) and V in normal years. Between 1987 and 2014, the FVC of mountain meadow showed a trend of decreasing, and alpine meadow increased firstly after declining and reached the peak in 2000, warm steppe, warm desert grassland were showed a contrary tendency, the FVC gradually improved after 2000.(3) The calculation was that average yield of fresh grass in alpine meadow, mountain meadow, warm steppe and warm desert grassland were 2092.71, 2864.74, 1456.13 and 1409.17 kg/hm2 respectively, total yield of fresh grass were 177300, 89900, 139900 and 142900 ton, which carrying capacity of 85000, 43100, 67100 and 68500 sheep unit respectively. Yield of in alpine meadow, mountain meadow, warm steppe and warm desert grassland between 3000~4500, 3000~6000,1500~3000 and 1500-3000 kg/hm2 respectively account for a large area.(4) Overall classification accuracy reached 90.21%,Kappa coefficient k=0.8735 by using decision tree classification, meet the basic requirements of overall classification accuracy ≥80%. Estimation of vegetation coverage and fresh grass productivity(not including temperate desert) of alpine meadow, mountain meadow, warm steppe, warm desert grassland and temperate desert by using improved dimidiate pixel model and regression model with NDVI and measured data, the model precision were more than 74% and 63% respectively, all possess better representative.
Keywords/Search Tags:Grassland Resources, Vegetation Coverage, Grassland productivity, Dynamic Change, Decision Tree Classification, Dimidiate Pixel Model
PDF Full Text Request
Related items