Font Size: a A A

Evaluation Of Grassland Degradation In West Jilin Province Based On Optical And Radar Images

Posted on:2019-04-25Degree:DoctorType:Dissertation
Country:ChinaCandidate:H YuFull Text:PDF
GTID:1363330569480930Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Grassland is an important ecosystem type and plays an irreplaceable role in maintaining regional sustainable development and ecological balance.Grassland degradation refers to the overall reduction in grassland productivity due to human activity and natural processes.It may be manifested as a reduction in the extent of grass cover,a reduction in the density of grass cover,a reduction in the output of forage material or as an increase in unpalatable grass species or even as a denudation of the underlying soil.Grassland degradation has been a hot topic in the study of grassland ecosystem.However,most of the studies on grassland degradation are based on grassland degradation area,grassland aboveground grass yield and species characteristics,but there are few studies of grassland degradation from the integrated on above ground grassland growth trend and underground grassland soil environment.Based on the region characteristics of west Jilin Province,this paper attempts to evaluate the grassland degradation comprehensively from the parameters of aboveground grassland(grassland fractional coverage for shot FVC and herbage yield for short HY)and parameters underground soil(electronic conductivity for short EC,pH).The comprehensive evaluation of grassland degradation made the results more close to the true situation of grassland degradation in west Jilin Province.According to the China's desertification classification indicator studies and China's national standard“parameters for degradation,sandification and salification of rangelands”(GB19377-2003)classification standard of grassland degradation,the degree of grassland degradation in west Jilin Province from 1990 to 2016 were estimated,which provides decision support and data support for ecological construction,resource optimization and sustainable development.Moreover,the result helps reverse the rangeland degradation and promote the rehabilitation of rangeland vegetation.The basic conclusions are drawn as follows:(1)Combined with radar(Sentinel-1A)and optical(Landsat TM/OLI)remote sensing data to map grassland distribution in 2016.Establishment of interpretation marks according to the optical image features of land cover types in western Jilin Province.This paper takes full advantages of object oriented segmentation and decision tree classification,to establish decision tree rule set by various types(vegetation and non-vegetation for optical image,waterbody and built-up land for radar imager)on optical and radar imageries.After that,the spatial information of grassland in 2016 is extracted from the interpretation of two types of images by comparing crossing and excluding.This method can get more high classification accuracy than only using optical images.Compared with the results of optical image interpretation alone,the accuracy of this method is higher.The classification accuracy increased from 88.78%to 90.10%.The distribution information of grassland in 1990,2000 and 2010 were obtained by change detection.The total accuracy of the three periods was 88.33%,87.95%and 89.10%,respectively.(2)The change of grassland area in west Jilin Province from 1990 to 2016 was analyzed,and the conversion relationship between grassland and other types in this area was grasped to provide data support for the comprehensive evaluation of grassland degradation in the later stage.The result shows that grassland area in the study area decreased first and then increased in the past 26 years.From 1990 to 2000,the total number of grassland decreased by 66,300 hm~2.The phenomenon of grassland degradation was obvious in this period,and a large number of grasslands were developed and utilized into dryland fields.Degradation region was mainly occurred in the north of the west Jilin Province.From 2000 to 2010,the grassland area increased by 38,300 hm~2.The increase area mainly transfer from wetland and saline-alkali bare land grassland,and the increasing area mainly distributed the central part of Da'an city and the east of Taonan city,but the grassland area still did not recover to the area of1990.From 2010 to 2016,the grassland area reduction rate is about half of the rate of increase from 2000 to 2010.The grassland decreased region was mainly concentrated in Da'an City and Zhenlai County.(3)According to the national grassland degradation standard(GB19377-2003),the grassland degradation parameters are determined;the relation models between the measured data of grassland degradation index and the spectrum of remote sensing image are established.In this paper,the above ground parameters(FVC,HY)and underground soil parameters(EC,pH)are used as the evaluation indexes of grassland degradation evaluation.Through correlation analysis and regression analysis to select the optimal regression models of grassland degradation parameters.The result indicated that NDVI showed the highest correlation on above ground parameters,and Green band,NDVI and salinity indes(SI)showed the higher correlation than other bands and spectral indexs on underground parameters.The optimal regression models for every grassland degradation parameters are established by regression analysis.In order to construct the historical grassland degradation parameters model more accurately,all spectral bands are normalized before the model inversion.The aim is to reduce the influence of different image reflectivity in different years and time on establishing inversion model to reduce the effect of different image reflectivity on model building.The prediction accuracy was assessed with 55 validating samples using R~2,which revealed R~2 of the FVC,HY,EC and pH is reaching 0.813,0.567,0.862 and 0.731,respectively.Based on the spatial distribution of the parameters of grassland degradation in different periods,it was found that the FVC of grassland in west Jilin Province decreased from 1990 to 2016.FVC changed from 0.45 in 1990 to0.29 in 2016.However,the coverage increased slightly(0.04%)from 2010 to 2016;HY also showed a downward trend,from 271.85g/m~2 in 1990 to 198.91 g/m ~2 in 2016,and the HY in west Jilin Province was almost equal from 2010 to 2016.EC is fluctuating during past 26 years,the EC value increased obviously from 1990 to 2000,the value of EC in 2000 was about twice that of 1990,then the value of EC decreased,and the value of EC almost returned to 1990 level in 2016.The change trend of p H is the same as EC,showing a trend of first increasing and then decreasing.The difference is that the pH of grassland in the west of Jilin Province did not return to1990 state in 2016,and the salinization degree of grassland in this area was significant from pH range of(8.71-9.76)and EC range of(0.45-1.40 mS/cm).(4)According to the criteria of the China's desertification classification indicator studies and China's national standard“parameters for degradation,sandification and salification of rangelands”(GB19377-2003)classification standard of grassland degradation,this paper calculated the comprehensive score for the evaluated grassland degradation grade with weighted summation method,and achieve the result of grassland degradation in different grades on spatial distribution.The forecast result shows that roughly three quarters of the study area has been affected by degradation to various levels of severity during 1990 to 2000.The total area of grassland degradation is 34.30 million hm~2,the grade of degradation is mainly severe degeneration which mainly occurred in the middle of the area in the low-lying areas of Da'an City and Gan'an County,the area is 22.96 million hm~2;During 2000-2010,the degraded area of grassland is 25.08 million hm~2,and slightly and moderately degraded are the mainly grade of grassland degradation in this period,area is 10.71million hm~2 and 10.05 million hm~2,respectively.During 2010 to 2016,grassland degradation area is 29.48 million hm~2,with severely degraded grassland area of 17.65million hm~2.In conclusion,the grassland degradation has controlled in west Jilin Province from 1990-2016,grassland degradation degree from severe degradation gradually into moderate degraded and slightly degraded in this period.(5)The degradation factors of grassland in the west of Jilin Province are closely related to the climatic environment with less rainfall and large evaporation,and the grassland degradation is obviously affected by the escalation of grass and livestock contradictions,extensive reclamation of farmland,the disturbance of human beings,etc.However,the implementation of ecological engineering after 2000,especially the alkali control project and grassland protection project,has controlled the rate of grassland degradation on a certain extent.
Keywords/Search Tags:Grass Degradation, Degradation parameters, Landsat, Sentinel-1A, West Jilin Province
PDF Full Text Request
Related items