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

Accueil du site → Doctorat → Allemagne → 2015 → Implementing an Agro-Environmental Information System (AEIS) Based on GIS, Remote Sensing, and Modelling — A case study for rice in the Sanjiang Plain, NE-China

Universität zu Köln (2015)

Implementing an Agro-Environmental Information System (AEIS) Based on GIS, Remote Sensing, and Modelling — A case study for rice in the Sanjiang Plain, NE-China

Zhao, Quanying

Titre : Implementing an Agro-Environmental Information System (AEIS) Based on GIS, Remote Sensing, and Modelling — A case study for rice in the Sanjiang Plain, NE-China

Auteur : Zhao, Quanying

Université de soutenance : Universität zu Köln.

Grade : Doktorgrades der Mathematisch-Naturwissenschaftlichen Fakultät 2015

Résumé
Information on agro-ecosystems is crucial for understanding the agricultural production and its impacts on the environment, especially over large agricultural areas. The Sanjiang Plain (SJP), covering an area of 108 829 km², is a critical food base located in NE-China. Rice, soya bean and maize are the major crops in the SJP which are sold as commercial grain throughout China. The aim of this study is to set up an Agro-Environmental Information System (AEIS) for the SJP by employing the technologies of geographic information systems (GIS), remote sensing (RS), and agro-ecosystem modelling. As the starting step, data carrying interdisciplinary information from multiple sources are organized and processed. For an AEIS, geospatial data have to be acquired, organized, operated, and even regenerated with good positioning conditions. Georeferencing of the multi-source data is mandatory. In this thesis, high spatial accuracy TerraSAR-X imagery was used as a reference for georeferencing raster satellite data and vector GIS topographic data. For the second step, the georeferenced multi-source data with high spatial accuracy were integrated and categorized using a knowledge-based classifier. Rice was analysed as an example crop. A rice area map was delineated based on a time series of three high resolution FORMOSAT-2 (FS-2) images and field observed GIS topographic data. Information on rice characteristics (i.e., biomass, leaf area index, plant nitrogen concentration and plant nitrogen uptake) was derived from the multi-temporal FS-2 images. Spatial variability of rice growing status on a within-field level was well detected. As the core part of the AEIS, an agro-ecosystem modelling was then applied and subsequently crops and the environmental factors (e.g., climate, soil, field management) are linked together through a series of biochemical functions inherent in the modelling. Consequently, the interactions between agriculture and the environment are better interpreted. In the AEIS for the SJP, the site-specific mode of the DeNitrification-DeComposition (DNDC) model was adapted on regional scales by a technical improvement for the source code. By running for each pixel of the model input raster files, the regional model assimilates raster data as model inputs automatically. In this study, detailed soil data, as well as the accurate field management data in terms of crop cultivation area (i.e. rice) were used as model inputs to drive the regional model. Based on the scenario optimized from field observation, rice yields over the Qixing Farm were estimated and the spatial variability was well detected. For comparison, rice yields were derived from multi-temporal FS-2 images and the spatial patterns were analysed. As representative environmental effects, greenhouse gas of nitrous oxide (N2O) and carbon dioxide (CO2) emitted from the paddy rice fields were estimated by the regional model. This research demonstrated that the AEIS is effective in providing information about (i) agriculture on the region, (ii) the impacts of agricultural practices on the environment, and (iii) simulation scenarios for sustainable strategies, especially for the regional areas (e.g. the SJP) that is lacking of geospatial data.

Présentation

Version intégrale (7 Mb)

Page publiée le 30 octobre 2017, mise à jour le 12 décembre 2017