Ryutaro Tateishi

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.

Mapping evapotranspiration and water balance for global land surfaces

Abstract Evapotranspiration returns approximately 60–80% of precipitation back to the atmosphere, where it becomes the source of future precipitation. Any variations in the distribution and availability of surface water will affect human activities and undoubtedly result in some socio-economic adjustments. New water demands for irrigation, as well as non-agricultural purposes, are growing with increased population and industrialization. It is reasonable to conclude that our knowledge of broad-scale evapotranspiration patterns is incomplete and that improved data sets are needed. Therefore, in this study, monthly global data sets of evapotranspiration and water balance were produced using a simplified water balance model and published global data sets. For validation of the global data sets, results were compared with information obtained by previous investigations that used independent data and analytical approaches.

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.

Remote Sensing of Fractional Green Vegetation Cover Using Spatially-Interpolated Endmembers

Fractional green vegetation cover (FVC) is a useful parameter for many environmental and climate-related applications. A common approach for estimating FVC involves the linear unmixing of two spectral endmembers in a remote sensing image; bare soil and green vegetation. The spectral properties of these two endmembers are typically determined based on field measurements, estimated using additional data sources (e.g., soil databases or land cover maps), or extracted directly from the imagery. Most FVC estimation approaches do not consider that the spectral properties of endmembers may vary across space. However, due to local differences in climate, soil type, vegetation species, etc., the spectral characteristics of soil and green vegetation may exhibit positive spatial autocorrelation. When this is the case, it may be useful to take these local variations into account for estimating FVC. In this study, spatial interpolation (Inverse Distance Weighting and Ordinary Kriging) was used to predict variations in the spectral characteristics of bare soil and green vegetation across space. When the spatially-interpolated values were used in place of scene-invariant endmember values to estimate FVC in an Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) image, the accuracy of FVC estimates increased, providing evidence that it may be useful to consider the effects of spatial autocorrelation for spectral mixture analysis.

Multiscale remote sensing data segmentation and post-segmentation change detection based on logical modeling: Theoretical exposition and experimental results for forestland cover change analysis

Quantification of forestland cover extents, changes and causes thereof are currently of regional and global research priority. Remote sensing data (RSD) play a significant role in this exercise. However, supervised classification-based forest mapping from RSD are limited by lack of ground-truth- and spectral-only-based methods. In this paper, first results of a methodology to detect change/no change based on unsupervised multiresolution image transformation are presented. The technique combines directional wavelet transformation texture and multispectral imagery in an anisotropic diffusion aggregation or segmentation algorithm. The segmentation algorithm was implemented in unsupervised self-organizing feature map neural network. Using Landsat TM (1986) and ETM+ (2001), logical-operations-based change detection results for part of Mau forest in Kenya are presented. An overall accuracy for change detection of 88.4%, corresponding to kappa of 0.8265, was obtained. The methodology is able to predict the change information a-posteriori as opposed to the conventional methods that require land cover classes a priori for change detection. Most importantly, the approach can be used to predict the existence, location and extent of disturbances within natural environmental systems.

Global mapping of urban built-up areas of year 2014 by combining MODIS multispectral data with VIIRS nighttime light data

ABSTRACT An improved methodology for the extraction and mapping of urban built-up areas at a global scale is presented in this study. The Moderate Resolution Imaging Spectroradiometer (MODIS)-based multispectral data were combined with the Visible Infrared Imager Radiometer Suite (VIIRS)-based nighttime light (NTL) data for robust extraction and mapping of urban built-up areas. The MODIS-based newly proposed Urban Built-up Index (UBI) was combined with NTL data, and the resulting Enhanced UBI (EUBI) was used as a single master image for global extraction of urban built-up areas. Due to higher variation of the EUBI with respect to geographical regions, a region-specific threshold approach was used to extract urban built-up areas. This research provided 500-m-resolution global urban built-up map of year 2014. The resulted map was compared with three existing moderate-resolution global maps and one high-resolution map in the United States. The comparative analysis demonstrated finer details of the urban built-up cover estimated by the resultant map.

Multitemporal comparative analysis of TRMM-3B42 satellite-estimated rainfall with surface gauge data at basin scales: daily, decadal and monthly evaluations

In order to examine the reliability and applicability of Tropical Rainfall Measuring Mission (TRMM) and Other Satellites Precipitation Product (3B42) Version 6 (TRMM-3B42) at basin scales, satellite rainfall estimates were compared with geostatistically interpolated reference data from 12 rain gauge stations for three consecutive years: 2005, 2006 and 2007. Gauge–TRMM-3B42 statistical properties for daily, decadal and monthly multitemporal precipitations were compared using the following cross-validation continuous statistical measures: mean bias error (MBE), root mean square difference (RMSD), mean absolute difference (MAD) and coefficient of determination (r 2) metrics. The averaged spatial–temporal comparisons showed that the TRMM-3B42 rainfall estimates were much closer to the geostatistically interpolated gauge data, with minimal biases of −0.40 mm day−1, −1.78 mm decad−1 and −6.72 mm month−1 being observed in 2006. In the same year, the gauge and TRMM-3B42 rainfall estimates marginally correlated better than in 2005 and 2007, with the daily, decadal and monthly coefficients of determination being 82.2%, 93.9% and 96.5%, respectively. The results showed that the correlations between the gauge-derived precipitation and the TRMM-3B42-derived precipitation increased with increasing temporal intervals for all three considered years. Quantitatively, the TRMM-3B42 observations slightly overestimated the precipitations during the wet seasons and underestimated the observed rainfall during the dry seasons. The results of the study show that the estimates from TRMM-3B42 precipitation retrievals can effectively be applied in the interpolation of missing gauge data, and in the verification of precipitation uncertainties at the basin scales with minor adjustments, depending on the timescales considered.

Agricultural land monitoring in the Egyptian Nile delta using Landsat data

The quest to bring new land under cultivation has been the cornerstone of Egyptian agricultural policy since the 1950s. At the same time, overpopulation in the Nile delta and valley, and the gradual shifting of economy from an agricultural to an industrial base together with different forms of desertification processes have led to the loss of highly fertile agricultural land. Therefore, there is an urgent need to address and analyse these changes in order to serve the development plans of the country. In this paper a maximum likelihood supervised classification was applied to subsets of the Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper (ETM +) images acquired on 26 March 1989 and 19 March 2001, respectively, to monitor agricultural land changes in the Eastern Nile Delta of Egypt between the two dates. A supervised classification was carried out with the six reflective bands for the two images individually and ground truth data were used to assess the accuracy of the classification. Generally, it was found that urban areas increased by 34%, which was considered to be the highest land cover change in the study area. This increase in urban areas took place on desert and highly productive agriculture land in the Nile delta as a result of overpopulation and economic growth. In spite of the continuous efforts to reclaim the desert and increase cultivated land, agricultural land has been decreased by about 2%. At the same time, more concern is required to prevent the urban encroachment at the expense of productive agricultural land. The results are analysed and discussed.

Using geographically weighted variables for image classification

In this study, geographically weighted variables calculated for two tree species, Cryptomeria japonica (Sugi) and Chamaecyparis obtusa (Hinoki), were used in addition to spectral information to classify the two species and one mixed forest class. Spectral values (digital numbers for each band) of ‘Sugi’ and ‘Hinoki’ training samples were used to predict the spectral values for the two species at other locations using the inverse distance weighting (IDW) interpolation method. Next, the similarity between each pixels spectral values and their IDW predicted values was calculated for both of the tree species. The similarity measures are considered to be geographically weighted because nearer training samples have more of an impact on their calculation. The use of geographically weighted variables resulted in an increase in overall accuracy from 82.2% to 85.9% and an increase in the kappa coefficient from 0.740 to 0.795 for a support vector machine classification.

Comparison of Landsat image classification methods for detecting mangrove forests in Sundarbans

Remote-sensing images taken from the Landsat Enhanced Thematic Mapper Plus (ETM+) sensor with a spatial resolution of 30 m were applied for mapping and inventory of mangrove forest areas in Sundarbans, on both sides of the border between Bangladesh and India. Three different classification methods – unsupervised classification with k-means clustering, supervised classification using the maximum likelihood decision rule, and band-ratio supervised classification – were tested and compared in terms of the top of the atmosphere reflectance images. Spectral signature and principal component analyses were applied to select the appropriate band combinations prior to the band ratio–supervised classification. Our results show that the band ratio method is superior to the unsupervised or supervised classification methods considering the visual inspection, producers and users accuracy, as well as the overall accuracy of the all the classes in the image. The best discrimination of mangrove/nonmangrove boundary can be achieved when the combinations of B4/B2 (band 4/band 2), B5/B7, and B7/B4 are employed from the ETM+ bands.