Publications

Kraatz, S; Khanbilvardi, R; Romanov, P (2016). River ice monitoring with MODIS: Application over Lower Susquehanna River. COLD REGIONS SCIENCE AND TECHNOLOGY, 131, 116-128.

Abstract
Spatially detailed characterization of the distribution, amount and timing of river ice is important for identifying and predicting potential ice hazards. Although information on the ice cover over inland water bodies is provided within MODIS snow products (MOD10 and MYD10), this information has little practical value for river ice monitoring. First, many rivers are not properly resolved in the MODIS land/water mask. Second, the cloud mask incorporated in the product is conservative and therefore results in reduced effective area coverage of the product. Third, the accuracy of the incorporated cloud mask (MOD35) depends on the particular setting and underlying land/water mask. The MODIS cloud mask is not suitable for this particular setting, if frequent ice observations are desired. In this study we present an alternative river ice monitoring algorithm for MODIS that identifies river ice both in cloud-free conditions and through some semitransparent clouds. As input the semitransparent cloud algorithm (STC) uses an improved land/water mask accurately delineating the river channel along with bands 4 and 7 data from the MODIS surface reflectance product. The algorithm was developed for the MODIS instrument and comprehensively tested with MODIS Aqua data over the Lower Susquehanna River for the 2014 winter season. It is shown that river ice cover retrievals made with STC are highly consistent with in situ observed ice processes, for both individual scenes and the overall winter period. Owing to a better-suited cloud masking algorithm, the new technique yields nearly twice as many usable river observations during the ice-bearing period as compared to the MODIS cloud cover product. The presented approach provides potential for more timely identification of river ice changes and hence more accurate prediction of ice-related hazards using MODIS data. (C) 2016 Elsevier B.V. All rights reserved.

DOI:
10.1016/j.coldregions.2016.09.012

ISSN:
0165-232X