02365nam a2200229 a 450000100080000000500110000800800410001910000160006024501540007626001180023030000180034852015960036665000120196265000150197465300320198965300260202170000150204770000220206270000180208470000180210270000150212019023172019-05-03       2011    bl uuuu     u00u1 u    #d1 aBEUCHLE, R.  aGlobal tropical forest cover change assessment with medium spatial stellite imagery using a systematic sample grid - data, methods and first results.  aIn: SIMPÓSIO BRASILEIRO DE SENSORIAMENTO REMOTO, 15., 2011, Curitiba. Anais... São José dos Campos: INPEc2011  ap. 2981-2988.  aAt the Joint Research Centre (JRC) of the European Commission, a methodology has been developed to monitor the pan-tropical forest cover with remote sensing data for the years 1990-2000-2005 in Latin America, Southeast Asia and Africa on the basis of over 4000 sample units sample units with a dimension of 20 km by 20 km located at every full latitude and longitude degree confluence. From the Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper (ETM) instruments, images with low cloud impact from the epochs around the years 1990, 2000 and 2005 were selected and subsets covering the sample units were cut-out, pre-processed, segmented and classified in five different land cover classes in order to build global and regional statistics on tropical forest cover change. The data was validated in three steps, internal correction of wrongly classified objects, external (national or regional) expert validation and internal harmonization of the data. In this paper, the data collection and the workflow of the forest cover change assessment for the epochs 1990 and 2000 is presented. Parts of the results for the Brazilian Amazon have been validated by comparing with interpretations of corresponding samples carried out by the Instituto Nacional de Pesquisas Espaciais (INPE), showing a very high correlation. Further, the figure produced by INPE through the PRODES program on gross deforestation for the years 1990-2000 was compared to the figure calculated on basis of the JRC results for the respective area, where the JRC estimate that was ca. 10% higher than the INPE estimate.  aLandsat  amonitoring  aObject-based classification  aTropical forest cover1 aEVA, H. D.1 aMIRANDA, E. E. de1 aHOLLER, W. A.1 aOSHIRO, O. T.1 aACHARD, F.