Laboratory for Remote Sensing Hydrology and Spatial Modeling
Professor Ke-Sheng Cheng (Email: rslab@ntu.edu.tw)
RSLAB_BSE_NTU
No. 1, Section 4, Roosevelt Road
Bioenvironmental Syst. Eng., National Taiwan University
Introduction
The Laboratory for Remote Sensing Hydrology and Spatial Modeling at the Department of Bioenvironmental Systems Engineering, National Taiwan University (RSLAB) focuses its research works on three inter-related fields: hydrological processes, environmental remote sensing and spatial modeling. Almost all hydrological and environmental processes exhibit heterogeneity and random variations in time and spatial domains. These processes often involve multi-scale phenomena, and thus stochastic modeling and simulation, in addition to fundamental physical principles, plays an essential role in modeling these complex processes. At RSLAB, we dedicate our research efforts to stochastic modeling of hydrological and environmental processes using field measurements and remote sensing images. Stochastic modeling and simulation enables us not only to monitor and/or forecast hydrological and environmental changes but also assess the risks and impacts of such changes.
Does urbanization increase diurnal land surface temperature variation? Evidence and implications. Landscape and Urban Planning, 2017.
On the criteria of model performance evaluation for real-time flood forecasting Stochastic Environmental Research and Risk Assessment, 2017.
A Feature Space Indicator Kriging Approach for Remote Sensing Image Classification. IEEE Transactions on Geoscience and Remote Sensing. Chiang, J.L., Liou, J.J., Wei, C., Cheng, K.S., 2014.
All-English Graduate Curriculum on Environmental Informatics and Risk Management (CEIRM)
Program overview
In light of its unique geographic and environmental characteristics, Taiwan is particularly vulnerable to natural disasters. Findings from many studies have demonstrated that climate changes can result in much greater impacts to island-countries like Taiwan than other continental countries. It is therefore of vital importance for both government agencies and engineering consultancy firms to rely on environmental data and information for sound engineering practices and decision making in disaster remediation, ecological restoration, soil and water conservation, adaptation to climate changes, etc. This comprehensive curriculum provides an integrated and multidisciplinary program that spans a very wide spectrum including environmental database building and analysis, environmental monitoring and instrumentation, landscape ecological measuring and planning, restoration ecology, spatiotemporal stochastic modeling and simulation, risk assessment and management, environmental system/process modeling, and climate change impact assessment.
The CEIRM emphasizes not only the fundamental theories and computer skills for environmental modeling and risk assessment, but also practical applications to solve real world problems. A few such applications include quantitative assessment of the climate change impact on urban inundation, drought indicators and drought forecasting, ecological indicators for river ecology restoration, evaluating the effect of urbanization on diurnal temperature variations, etc.
Featured Research
Assessing Uncertainty in LULC Classification Accuracy by Using Bootstrap Resampling Remote Sensing, 2016. L.H. Hsiao and K.S. Cheng.
Supervised land-use/land-cover (LULC) classifications are typically conducted using class assignment rules derived from a set of multiclass training samples. Consequently, classification accuracy varies with the training data set and is thus associated with uncertainty. In this study,we propose a bootstrap resampling and reclassification approach that can be applied for assessing not only the uncertainty in classification results of the bootstrap-training data sets, but also the classification uncertainty of individual pixels in the study area. Two measures of pixel-specific classification uncertainty, namely the maximum class probability and Shannon entropy, were derived from the class probability vector of individual pixels and used for the identification of unclassifiedpixels. Unclassified pixels that are identified using the traditional chi-square threshold technique represent outliers of individual LULC classes, but they are not necessarily associated with higher classification uncertainty. By contrast, unclassified pixels identified using the equal-likelihood technique are associated with higher classification uncertainty and they mostly occur on or near the borders of different land-cover.
RSLAB - NTU
Prof. Ke-Sheng Cheng
RSLAB_BSE_NTU
No. 1, Section 4, Roosevelt Road
Bioenvironmental Syst. Eng., National Taiwan University