Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Chine → 2016 → Identification and Assessment of Land Degradation Using Remote Sensing Techniques

China Forestry Science Academe (2016)

Identification and Assessment of Land Degradation Using Remote Sensing Techniques

孙斌;

Titre : Using Remote Sensing Techniques

Auteur : 孙斌;

Grade : Doctoral Dissertation 2016

University : China Forestry Science Academe Site Web

Résumé
Under the adverse effects of climate change and human activity, land degradation has been recognized as one of the major threats to global environment impacting directly on human well-being and food security, it may eventually influence the human living and development. With the development of earth observation technology and enrichment of remote sensing datas, remote sensing has become an important tool for regional land degradation assessment. Long time series of vegetation parameters have become available based on remote sensing, thus degraded land could be identified and assessed by analyzing the changes of the time series vegetation index. However, time series vegetation index will fluctuate severely due to the impact of climate change, especially the fluctuation of annual precipitation, thereby the land production capacity could not be determined accurately. Meanwhile, numerous mixed evaluation index, unclear assessment "benchmark", different standards of land degradation degree division and unsuitable for repeated scanning were all hindered the development of land degradation assessment based on remote sensing.Therefore, in this paper, Inner Mongolia Autonomous Region was selected as the study area. Based on the time series MODIS NDVI data during 2001 to 2012 and reference the existing regional land degradation assessment method, a technique system for land degradation assessment using remote sensing in Inner Mongolia was developed, which will be characterized by definition of land degradation, suitable for use at large-scale, sustained by remote sensing technology and easy to operate. The key techniques in land degradation assessment technique system “degraded lands area”, “land degradation degree” and “driving factor” were studied. Then, the spatial patterns and dynamics of degraded lands in recent 12 years and driving process of land degradation in Inner Mongolia were be analyzed. Finally, the results of the land degradation assessment were validated tested and verified by the relevant collected data. This research will provide technical support and scientific reference data for land degradation assessment and management in Inner Mongolia, even in the whole northern China.The main conclusions are drawn as follows :(1) Based on the 34 years meteorological data, the scopes of background climate zones of land degradation in Inner Mongolia were determined. The area of the general potential extend of desertification(PED) in Inner Mongolia is 7.58×105km2, 64.1% of study area’s terrestrial. By using Sen+Mann-Kendall and variety of mathematical statistics methods, the dynamic and variation of annual precipitation, annual temperature and moisture index(MI) were analysed in two period, 34 years and 12 years respectively. Results showed that, except the eastern part, the climatic dry/wet change was alternate and fluctuant in vast study area during 1979 to 2012, while under the influence of increased precipitation and decreased temperature, the main climatic dry/wet change in recent 12 years was wet.(2) Monthly and annual net primary productivity(NPP) was estimated over study during 2001 to 2012, by applying the improved Carnegie-Ames-Stanford Approach model(CASA) model, based on the monthly MODIS NDVI and contemporaneous meteorological data. Compared with the actual NPP data, there was a significant correlation between actual NPP and estimated NPP, R2=0.6029,RMSE=30.6. Results of the correlation analysis between annual and the main meteorological factors showed that, in our study areas, precipitation and accumulated temperature were the main limited meteorological factor for grassland, especially in arid region, semi-arid region and dry sub-humid region. While, accumulated temperature and solar radiation were the main limited meteorological factor for forestland, the influence of precipitation is no longer important. A new vegetation indictor——Moisture index net primary productivity(MNPP) was defined. By analyzing the response relationship between NPP, MNPP and MI, results showed that in most area of arid region, semi-arid region and dry sub-humid region, NPP has a significant positive correlation with MI. However, in humid region, the positive correlation was no longer significant positive correlation accompanied the increased MI. Meanwhile, the correlation between MNPP and MI was significant vegetative, accounted for 80.3% of the total study area.(3) According to the analysis of change trend of annual NPP and MNPP, coupling Relationship between NPP, MNPP and MI at pixel scale, 7 types of the vegetation change mode in degraded and non-degraded land were established. A new remote sensing technique system for identifying areas of degraded lands at regional scale was proposed.(4) According a new proposed NPP normalization method, the “relative benchmark” for severity assessment of land degradation was established. Based on the annual NPP and actual measured soil organic matter(SOM), the coupling characteristics between and vegetation were analyzed and the severity assessment of land degradation threshold of different landuse types was determined.(5) Based on the technique system for land degradation assessment using remote sensing we proposed above, the spatial patterns and dynamics of degraded lands in recent 12 years in Inner Mongolia were analyzed. Results showed that, from 2001 to 2012, the study area ecosystem experienced a restored process overall and particle areas existed the phenomenon of land degradation. The area of degraded lands and restored lands was 40685km2 and 120260km2, and account for 3.5% and 10.2% of Inner Mongolia’s terrestrial respectively. Among them, areas of grassland degradation and restoration were 28972km2 and 60924km2 respectively, and were the largest in all landuse type. The degraded lands have experienced an obviously process of stable to significant degraded. In the first 6 years, land degradation degree was slightly in vast degraded lands, but in the nearly 6 years, the situation was severe and the overall areas of severely degradation lands have a significant increase. The driving factors of land degradation and restoration in this area were distinguished over the past 12 years through multiple and partial regression methods, the main driver of land degradation in study area was identified as human activity and both human activity and climate change. The effects of both human activity and climate change were the main drivers of land restoration ; the single effect of human activity also played an important role in land restoration, whereas climate change had a small influence in land degradation and restoration.The innovations of this study are in the following two aspects :(1) A remote sensing technique for identifying areas of degraded lands at regional scale was developed. Combined with the new defined vegetation indictor MNPP, the coupling relationship of vegetation——climate in degraded lands were analyzed. The vegetation change mode of NPP and MNPP in degraded and non-degraded land were established, and then the technique for identifying areas of degraded lands was proposed. It solves the problem that the accurate scope and definition of land degradation is difficult in the past researches, especially in the large scale.(2) A new “Relative benchmark” for assessment of land degradation were achieved. A new NPP normalization method was proposed for the fist time. Based on the coupling characteristics of annual NPP and actual measured SOM, the threshold for severity assessment of land degradation of different landuse types was determined. It better solved the problem that lacking of “benchmark” in the former studies of land degradation assessment

Mots clés : Land degradation; Ecological remote sensing; Inner Mongolia Autonomous Region; Net Primary Production(NPP); Climate change; Land degradation degree; Land degradation driving forces;

Présentation (CNKI)

Page publiée le 18 janvier 2017, mise à jour le 11 septembre 2017