Nguyen Thanh Hoan

Production of global land cover data GLCNMO

Abstract Global land cover is one of the fundamental contents of Digital Earth. The Global Mapping project coordinated by the International Steering Committee for Global Mapping has produced a 1-km global land cover dataset – Global Land Cover by National Mapping Organizations. It has 20 land cover classes defined using the Land Cover Classification System. Of them, 14 classes were derived using supervised classification. The remaining six were classified independently: urban, tree open, mangrove, wetland, snow/ice, and water. Primary source data of this land cover mapping were eight periods of 16-day composite 7-band 1-km MODIS data of 2003. Training data for supervised classification were collected using Landsat images, MODIS NDVI seasonal change patterns, Google Earth, Virtual Earth, existing regional maps, and experts comments. The overall accuracy is 76.5% and the overall accuracy with the weight of the mapped area coverage is 81.2%. The data are available from the Global Mapping project website (http://www.iscgm.org/). The MODIS data used, land cover training data, and a list of existing regional maps are also available from the CEReS website. This mapping attempt demonstrates that training/validation data accumulation from different mapping projects must be promoted to support future global land cover mapping.

A hybrid pansharpening approach and multiscale object-based image analysis for mapping diseased pine and oak trees

We developed a multiscale object-based classification method for detecting diseased trees (Japanese Oak Wilt and Japanese Pine Wilt) in high-resolution multispectral satellite imagery. The proposed method involved (1) a hybrid intensity–hue–saturation smoothing filter-based intensity modulation (IHS-SFIM) pansharpening approach to obtain more spatially and spectrally accurate image segments; (2) synthetically oversampling the training data of the ‘Diseased tree’ class using the Synthetic Minority Over-sampling Technique (SMOTE); and (3) using a multiscale object-based image classification approach. Using the proposed method, we were able to map diseased trees in the study area with a users accuracy of 96.6% and a producers accuracy of 92.5%. For comparison, the diseased trees were mapped at a users accuracy of 84.0% and a producers accuracy of 70.1% when IHS pansharpening was used alone and a single-scale classification approach was implemented without oversampling the ‘Diseased tree’ class.

Satellite Image Pansharpening Using a Hybrid Approach for Object-Based Image Analysis

Intensity-Hue-Saturation (IHS), Brovey Transform (BT), and Smoothing-Filter-Based-Intensity Modulation (SFIM) algorithms were used to pansharpen GeoEye-1 imagery. The pansharpened images were then segmented in Berkeley Image Seg using a wide range of segmentation parameters, and the spatial and spectral accuracy of image segments was measured. We found that pansharpening algorithms that preserve more of the spatial information of the higher resolution panchromatic image band (i.e., IHS and BT) led to more spatially-accurate segmentations, while pansharpening algorithms that minimize the distortion of spectral information of the lower resolution multispectral image bands (i.e., SFIM) led to more spectrally-accurate image segments. Based on these findings, we developed a new IHS-SFIM combination approach, specifically for object-based image analysis (OBIA), which combined the better spatial information of IHS and the more accurate spectral information of SFIM to produce image segments with very high spatial and spectral accuracy.

Tropical forest mapping using a combination of optical and microwave data of ALOS

It is difficult to monitor forests in tropical regions with frequent cloud cover using optical remote-sensing data. Adequate multi-temporal, high-resolution imagery is often not available. Microwave imagery is able to penetrate cloud cover, enabling imagery of the land surface to be recorded more frequently. This study seeks to improve tropical forest mapping by combining optical and microwave imagery, with one of the main objectives being the discrimination of planted and natural forests. First, multi-spectral Advanced Land Observing Satellite (ALOS)/Advanced Visible and Near Infrared Radiometer type 2 (AVNIR-2) images were used to create a forest and land-cover classification of the study area. Subsequently, ALOS/Phased Array type L-band Synthetic Aperture Radar (PALSAR) single-polarized and dual-polarized microwave images were used to generate forest and land-cover masks to be used in combination with the ALOS/AVNIR-2 classification. The overall accuracy of the ALOS/AVNIR-2 classification was 77%. When the ALOS/PALSAR masks were used in combination with the ALOS/AVNIR-2 classification, the overall accuracy increased to 88% with higher than 90% accuracy for the main forest classes.

Global urban characterization using population density, DMSP and MODIS data

This is study of urban characterization from coarse resolution satellite images. Urban environments are so heterogeneous. It is necessary to simplify them as combination of basic land cover materials in order to enable quantitative studies. A comparative analysis of three-component features (Population density, NDVI, DMSP) for 25 urban areas worldwide provides a physical basis of global urban characterization. The result of analysis indicate that the characteristics of there cities can be accurately described as linear combinations of population density, DMSP and vegetations end members within a two dimensional mixing space containing over 80% of variance in the observed data. Global urban characterizations provide a basis of mapping the spatial extent of human settlements using coarse resolution satellite imagery.

New urban map of Eurasia using MODIS and multi-source geospatial data

Abstract Urban areas are of paramount significance to both the individuals and communities at local and regional scales. However, the rapid growth of urban areas exerts effects on climate, biodiversity, hydrology, and natural ecosystems worldwide. Therefore, regular and up-to-date information related to urban extent is necessary to monitor the impacts of urban areas at local, regional, and potentially global scales. This study presents a new urban map of Eurasia at 500 m resolution using multi-source geospatial data, including Moderate Resolution Imaging Spectroradiometer (MODIS) data of 2013, population density of 2012, the Defense Meteorological Satellite Program’s Operational Linescan System (DMSP-OLS) nighttime lights of 2012, and constructed Impervious Surface Area (ISA) data of 2010. The Eurasian urban map was created using the threshold method for these data, combined with references of fine resolution Landsat and Google Earth imagery. The resultant map was compared with nine global urban maps and was validated using random sampling method. Results of the accuracy assessment showed high overall accuracy of the new urban map of 94%. This urban map is one product of the 20 land cover classes of the next version of Global Land Cover by National Mapping Organizations.

Global water mapping using MODIS tasseled cap indexes

Monitoring changes in water resources is very important due to waters importance for daily life. MODIS data is acquired at a high temporal resolution, and it is expected to continue acquiring data for many more years, making its good data source for monitoring water change at the continental to global scale. Existing global maps of water are not consistent with each other, and may not be suitable for water monitoring. The objective of this study is to develop a method for automated water mapping and monitoring at continental to global scales using MODIS data. MODIS 2008 data, a cloud free dataset processed by CEReS Chiba University, were used. MODIS tasseled cap indexes were calculated and integrated in some combination models with the reference of some existing maps and a DEM to map global water coverage. The water map produced in this study was compared with some other existing global maps such as GLOBCover, MOD44W, and GLCNMO 2003. Results of the comparisons show that product of this study, MODIS2008_WM, is better than other existing maps for the purpose of water mapping.

Improving tropical forest mapping using combination of optical and microwave data of ALOS

In the tropical regions, due to frequent cloud cover, optical remote sensing usually does not have adequate multi-temporal high resolution imagery to monitor phenology of forest. This paper seeks to improve mapping of tropical forest by combination of optical and microwave imagery. The study area is located in the southern part of Vietnam. Firstly, ALOS/AVNIR-2 images were used to create a forest map of the study area. Then, ALOS/PALSAR single-polarized and dual-polarized images were used to generate a second forest map of the study area. Discrimination of Planted Forest and Natural Forest is one of the most important purposes of this study. The overall accuracy of ALOS/AVNIR-2 classification result is 77.0%, while after combining with ALOS/PALSAR, it is increased up to 88.2%. The accuracy is higher than 90% for main forest classes.