Map Viewer
Data & Products
Services
Trace Gases
Aerosols
Clouds
Solar Radiation
Surface Parameters
Bio-Energy
Meteorology
Dynamics
Spectroscopy Data
Missions & Sensors
About WDC
User Account
Contact

NDMC
WDC Home









Local Weather
Oberpfaffenhofen

Land Cover Classfication

Description

The "MERIS Land Cover Classification of Germany (M_LCC_germany)" is a map product derived from multi-temporal, full-resolution MERIS data. For Germany, land cover map is a fixed grid map in Gauss-Kruger projection (zone 3) covering app. 695km x 895km with spatial resolution of 300m. The total size of the final map is 2317 samples by 2982 lines. The German land use is classified by applying the Land Cover Classification System (LCCS) resulting in a legend with 8 classes comparable with classes of the Global Land Cover 2000 (GLC2000).

Product Example:

Example of Land Cover Classification (LCC) of Germany for August 2004 derived from multi-spectral, multi-temporal satellite data (MERIS).

Methodology:

Within the MERIS Applications and Regional Products Project (MAPP), an operational algorithm was developed for calculating a map of land cover / land use of Germany, based on MERIS full resolution data. This algorithm is composed of a multi-spectral maximum likelihood classification and a multi-temporal analysis. By building a multi-spectral database and multi-temporal reference vectors, the classification can be updated automatically every year.

Within the scope of this work, an appropriate classification system (GLC 2000) and an applicable reference data set for validation (CORINE 2000) are proposed. The thematic legend and spatial resolution of the CORINE data are adapted to the GLC2000 legend and the MERIS spatial resolution. The spectral signatures of the eight considered land cover / land use classes are analysed and suitable test sites are defined. In this context, the occurrence of BRDF-effects in MERIS data is studied. Additional investigations concentrate on how the phenology of all considered classes is expressed in MERIS data and how this information should be taken into account in the multi-temporal analysis. On that basis, recommendations for minimising the computing time by channel reduction and for improving cloud identification are made. The results of the preparation of the MAPP classification are verified by means of a mono-temporal classification with the image analysis software ENVI 4.0. The resulting land cover / land use map for the beginning of August 2004 shows good agreement with the reference data set, concerning both coarse and detailed structures. The overall accuracy of the mono-temporal classification result is 57.3 %.

Precursor service:

Prior to the launch of MERIS, the LCC algorithm was tested using data from the “Modular Optoelectronic Scanner (MOS)”, flown on the Indian Remote Sensing platform IRS-P3 in order to assess its accuracy and precision. MOS data are comparable with MERIS data due to the spectral similarity of the sensor, their orbits, and the spatial resolution (~600m for MOS and ~300m for MERIS). The main difference is the limited swath width of MOS (~ 180km) compared to 1150km for MERIS.

Major users:

Scientific Organisations

Publications:

Arndt M., Günther K. P., Maier S. W.: Deriving Land Cover Information from multi-temporal MOS Data, In: 4th Berlin Workshop on Ocean Remote Sensing, “5 years of MOS-IRS”, Berlin, May 30th-June 1st, 2001, Ed.: Remote Sensing Technology Institute, DLR, Wissenschaft und Technik Verlag, 205 – 215, 2001.

Geßner, U.: Landbedeckungs- / Landnutzungsklassifizierung von Deutschland auf der Grundlage von Daten des Sensors MERIS, Thesis University of Augsburg, 113 pages, 2005. (download pdf (6,6MB))

Relevance

There is an increasing need for a precise description and classification of regional, continental and even global land cover and land use in order to deliver data that are needed for inventory purposes, urban development as well as pollution monitoring and air quality forecast. LCC data are frequently used in combination with models as e.g. biogenic volatile organic carbon emission models or tropospheric PM models.

Data Access

Contact: Dr. Kurt P. G√ľnther