Holistic Reduction to Compare and Create New Indices for Global Inter-Seasonal Monitoring: Case Study for High Resolution Surface Water Mapping

Abstract

A remote sensing method that integrates virtual sampling from formalized visual interpretations is proposed to facilitate land cover mapping and enhance its accuracy, with an emphasis on spatial and temporal scalability. Indices are widely used for mapping and monitoring surface water across space and time; however, they typically display some kind of limitation across different environments and seasons. A decision matrix framework based on observations derived from interpretation keys was designed to compare the performance of existing indices alongside a set of newly developed indices. This comparison helped to shortlist indices that warranted further evaluation and accuracy assessment to identify effective indices for global inter-seasonal surface water extent mapping. Additional visual inspections were conducted for criteria that remained unresolved by the decision matrix to examine index consistency across the seasons in a wide range of geographic settings around the world, and further reduce the shortlist. An accuracy assessment was performed for three new shortlisted indices. On a global scale, CAWI (Comprehensive Automatic Water Index) was the best-performing index. Its distinct binary data distribution provides the possibility of regional automatic Otsu thresholding. CAWI was determined to be compatible for Sentinel-2 and Landsat 8 sensors, providing the highest possible spatial resolution as well as the longest time series for retrospective analyses with freely available multispectral imagery. Two alternative indices were identified for sensors limited to the visible and NIR bands. The first index, CATWIC (Clear and Turbid Water Index Combination), split the classification of water into two components, with one index for generally clear water and another index for turbid water. The second, NDCHRWI (Normalized Difference Colourimetric High Resolution Water Index), applied the hue angle from a normalized difference RGB. Masking indices based on modified HSV Saturation equations were developed to reduce misclassification due to other high reflectance features. The indices’ overall accuracies, respectively, were: 94.97%, 94.51%, and 94.85%. This study concludes with recommendations for the application of different indices for sensors possessing shortwave infrared bands and for sensors limited to the visible and NIR bands, with a simple stratification of six zones for Global Surface Water monitoring.

High Resolution Forest Masking for Seasonal Monitoring with a Regionalized and Colourimetrically Assisted Chorologic Typology

“…The proposed chorologic typology has built upon previous studies that presented applications of colourimetry [60] and holistic reduction [81] to remote sensing. The results were modestly improved with an ecoregional stratification, which reduced the multispectral variation in land cover features across the full extent of the study area.…”

“…Holistic reduction addresses the need to balance holism and reductionism in ecological studies in order to explain outcomes by looking at parts of complex systems against the desire to understand how the parts work together in a fully functioning system [80]. A holistic reductionist approach to remote sensing [60,81] was undertaken to reduce analytical complexity by interpreting a single reduced class with an emphasis on large-scale performance, scalability, and global transferability.…”

“…The development of the candidate indices was guided by a holistic visual assessment of the omission or commission of the colourimetric benchmarks specified in the forest and non-forest SII keys with a decision matrix as a process for refinement [60,81]. For every seasonal image composite derived from Sentinel-2 data, each index was individually analysed and evaluated in a GIS at 100 m resolution.…”

“…Forest masking requires at least a pair of indices to approximate separability, and the selection of accompanying indices will vary depending on ecoregions [86]. Colourimetrically assisted classification has been shown to facilitate interpretation and communicability and significantly reduce (or even eliminate) sampling effort [60,81]. Colourimetric indices have therefore been proposed for three environmental indicators: (1) insolation, (2) water extent and (3) surface NIR saturation.…”

“…A previous study on water extent mapping determined that the hue and a modified saturation from an HSV transformation of a normalised difference RGB provided an effective solution to high-resolution water mapping for sensors limited to visible and NIR bands [81]. It was shown to be capable of classifying the full range of water types from clear to turbid in a single index without misclassifying vegetation as water:…”