Youngjoo Kwak

Prompt Proxy Mapping of Flood Damaged Rice Fields Using MODIS-Derived Indices

Flood mapping, particularly hazard and risk mapping, is an imperative process and a fundamental part of emergency response and risk management. This paper aims to produce a flood risk proxy map of damaged rice fields over the whole of Bangladesh, where monsoon river floods are dominant and frequent, affecting over 80% of the total population. This proxy risk map was developed to meet the request of the government on a national level. This study represents a rapid, straightforward methodology for estimating rice-crop damage in flood areas of Bangladesh during the large flood from July to September 2007, despite the lack of primary data. We improved a water detection algorithm to achieve a better discrimination capacity to discern flood areas by using a modified land surface water index (MLSWI). Then, rice fields were estimated utilizing a hybrid rice field map from land-cover classification and MODIS-derived indices, such as the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI). The results showed that the developed method is capable of providing instant, comprehensive, nationwide mapping of flood risks, such as rice field damage. The detected flood areas and damaged rice fields during the 2007 flood were verified by comparing them with the Advanced Land Observing Satellite (ALOS) AVNIR-2 images (a 10 m spatial resolution) and in situ field survey data with moderate agreement (K = 0.57).

Estimating floodwater from MODIS time series and SRTM DEM data

Real-time flood mapping with an automatic flood-detection technique is important in emergency response efforts. However, current mapping technology still has limitations in accurately expressing information on flood areas such as inundation depth and extent. For this reason, the authors attempt to improve a floodwater detection method with a simple algorithm for a better discrimination capacity to discern flood areas from turbid floodwater, mixed vegetation areas, snow, and clouds. The purpose of this study was to estimate a flood area based on the spatial distribution of a nationwide flood from the Moderate Resolution Imaging Spectroradiometer (MODIS) time series images (8-day composites, MOD09A1, 500-m resolution) and a digital elevation model (DEM). The results showed the superiority of the developed method in providing instant, accurate flood mapping by using two algorithms, which modified land surface water index from MODIS image and eight-direction tracking algorithm based on DEM data.

Near Real-Time Flood Volume Estimation From MODIS Time-Series Imagery in the Indus River Basin

Satellite images have been widely applied in near real-time flood inundation maps in many cases. Such images have significant potential to predict the time, place and scale of a flood event, and can be very useful in emergency response efforts. The detection of floodwaters and the estimation of flood volumes are important to determine a hazard in flood risk. In this study, we conducted surface water detection based on the spatial distribution of the 2010 Indus River flood, which affected the entire Pakistan area. A modified surface water index derived from near-real-time Moderate Resolution Imaging Spectrometer (MODIS) images coupled with a digital elevation model (DEM) was used. We also developed and applied a simplified algorithm to extract the 3D volume of floodplain surface water considering surface heights. The results found that the MODIS-DEM combined approach was feasible for automatic, instant flood detection. This approach shows a methodological possibility as an integrated algorithm for producing flood maps at local to global scales.

Estimation of flood volume in Chao Phraya River basin, Thailand, from MODIS images couppled with flood inundation level

River inundation satellite images are restricted to make real-time flood inundation maps in many cases. However, such images have significant potential to predict the time, place and scale of a flooding event, and can be very useful in emergency response efforts. The estimation of water extent boundary and flood volume is important to determine a fundamental hazard in flood risk assessment. In this study, an attempt was made to detect surface water in a severe flood event (the 2011 Thai flood) by applying modified remote sensing indices to near-real-time MODIS images. Flood volumes were also calculated for detected flood areas by using a proposed flood inundation level (FIL) model with the Digital Elevation Model (DEM). FILs were verified through field investigation. The results show that the MODIS-FIL combined approach is feasible for automatic, instant flooding detection.

Rapid Damage Assessment of Rice Crop After Large-Scale Flood in the Cambodian Floodplain Using Temporal Spatial Data

The objective of this study was to estimate rice crop damage over the entire Cambodia during a large flood event from July to November 2011. An integrated approach was applied to detect and monitor flood areas with flood depth and duration for near real-time rice crop damage estimation in 2011 by using MODIS time-series imagery. The combined data consists of developed MLSWI, EVI from MODIS, new FID from DEM, land use, and simplified empirical damage curves. These data are expected to play an important role in emergency response efforts and rapid risk assessment for high-risk flood areas in the Cambodian floodplain. A rice crop damage map will be generated, showing areas with different damage levels based on flood duration and floodwater depth, including 25% (8 days, below 1.5 m), 50% (8 days, over 1.5 m; 16 days, below 1.5 m), and 100% (16 days, over 1.5 m). The resulting map was validated and shows about 80% consistency with the government census based on field-scale investigation and survey.

Nation-wide flood risk assessment using inundation level model and MODIS time-series imagery

Real-time flood inundation mapping can be useful in emergency response efforts but is limited in capturing a wide area for assessment. In this paper, the authors suggest an improved extraction method of surface water with a simplified algorithm. The purpose of this study was to accurately estimate a flood area based on the spatial distribution of a nation-wide flood using a flood inundation level (FIL) model and a modified LSWI (MLSWI) from near real-time images from MODIS time series. This flood risk assessment study focused on reducing the susceptibility of nation-wide risk damage and improved the accuracy of extracting a vulnerable inundation area.

A new approach for rapid urban flood mapping using ALOS-2/PALSAR-2 in 2015 Kinu River Flood, Japan

A near-real-time algorithm to process satellite synthetic aperture radar (SAR) imagery for urban flood detection still remains a challenge. In this paper, we developed a new floodwater detection algorithm based on SAR backscattered intensity using pre- and post-flood data observed by the Japan Aerospace Exploration Agencys Advanced Land Observing Satellite-2, which provided very high-spatial-resolution L-band SAR images (PALSAR-2 SM1, 2.5 m spatial resolution) in the case of the Kinu River flood. With this algorithm, we first classified land use types in order to analyze the characterization of urban land surface backscatters. Change detection was then conducted for the pilot area using correlation coefficients based on differences in the magnitude of the backscattering intensity derived from building orientation before and right after the flood. In particular, the flooded building was more distinguishable than with a traditional detection method.

A determination of the earthquake disaster area by object-based analysis using a single satellite image

The final goal of this study is to create data to support emergency efforts in a disaster affected area by locating damaged buildings shortly after the disaster. In this study, prioritizing the practicality of the method for emergency purposes, we designed a method only to use a single satellite image of an affected area, eliminating the use of complex algorithms and auxiliary data. The uniqueness of our method lies in the application of an object-based region segmentation to images and the use of features of objects obtained from texture, hierarchical and other information in order to extract damaged buildings. Out of 26 features resulting from the analysis of objects, we found one feature and three combinations of two different features that are effective in extracting damaged buildings, such as Rectangular fit, Homogeneity, Number of sub-objects/Area, and Length of longest of edge/Area.

Large flood mapping using syncro water index coupling with hydro data and time series modis images

In this research, we proposed the Synchro Water Index (SWI) to detect widespread inundation extent in a transboundary river basin using the time-series Moderate Resolution Imaging Spectrometer (MODIS) data, a major contributor to progress in international flood monitoring. After removing clouds using the White-object Index (WoI), the multi-temporal processing coupled with in-situ water level data was applied to the 2015 Bangladesh flood for near-real-time nationwide rapid flood monitoring. The preliminary results showed that the maximum inundation area from SWI was underestimated to be smaller than the area from the solo use of modified land surface index, or 32% (29,900 km2) of the total area of Bangladesh.

Disaster risk reduction using image fusion of optical and SAR data before and after tsunami

This study applied supervised change detection to identify and estimate damaged urban surface conditions before and after a tsunami event in order to provide more accurate information for the implementation and enhancement of disaster risk reduction policies and strategies. Advanced remote sensing is crucial to support cost-effective emergency response activities for disaster risk assessment and management. This preliminary study, as an effort to propose a good case study in risk management, suggested that three main steps, i.e., filtering, fusing and classifying, should be adopted to perform change detection before and after a natural disaster. We fused very high-spatial-resolution multi-temporal optical images (0.6 m spatial resolution) and X-band SAR images (2.5 m spatial resolution). The study also revealed that the decision-level method, i.e. morphological transform, was the most promising in image fusion of filtered images to classify urban surfaces in tsunami damage assessment, compared with the pixel-level method, i.e. wavelet transform, and feature-level method, i.e. segmentation extraction. This paper reports, coupled with the results from the image fusion, that the preliminary results are good enough to obtain urban impervious surface estimation of a wide disaster risk area but not good enough to make clear amplitude images to identify individual buildings of dwelling zone. The proposed normalized change index (NCI), an important indicator for detecting changes, was found capable of providing better estimation of urban impervious surfaces, such as transport-related land (e.g., bridges and parking lots) and building roof tops in residential and industrial areas over a coastal zone in Rikuzen-takada City, devastated in the 2011 Great East Japan Earthquake.