Loading…

Improving the accuracy of MODIS 8-day snow products with in situ temperature and precipitation data

•We developed a methodology to optimize the accuracy of MODIS snow products.•Two meteorological filters can efficiently reduce snow overestimation error.•Cloud coverage on MODIS snow data also declined significantly. MODIS snow data are appropriate for a wide range of eco-hydrological studies and ap...

Full description

Saved in:
Bibliographic Details
Published in:Journal of hydrology (Amsterdam) 2016-03, Vol.534, p.466-477
Main Authors: Dong, Chunyu, Menzel, Lucas
Format: Article
Language:English
Subjects:
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•We developed a methodology to optimize the accuracy of MODIS snow products.•Two meteorological filters can efficiently reduce snow overestimation error.•Cloud coverage on MODIS snow data also declined significantly. MODIS snow data are appropriate for a wide range of eco-hydrological studies and applications in the fields of snow-related hazards, early warning systems and water resources management. However, the high spatio-temporal resolution of the remotely sensed data is often biased by snow misclassifications, and cloud cover frequently limits the availability of the MODIS-based snow cover information. In this study, we applied a four-step methodology that aims to optimize the accuracy of MODIS snow data. To reduce the cloud fraction, 8-day MODIS data from both the Aqua and Terra satellites were combined. Neighborhood analysis was applied as well for this purpose, and it also contributed to the retrieval of some omitted snow. Two meteorological filters were then applied to combine information from station-based measurements of minimum ground temperature, precipitation and air temperature. This procedure helped to reduce the overestimation of snow cover. To test this technique, the methodology was applied to the Rhineland-Palatinate region in southwestern Germany (approximately 20,000km2), where cloud cover is especially high during winter and surface heterogeneity is complex. The results show that mean annual cloud coverage (reference period 2002–2013) of the 8-day MODIS snow maps could be reduced using this methodology from approximately 14% to 4.5%. During the snow season, obstruction by clouds could be reduced by even a higher degree, but still remains at about 11%. Further, the overall snow overestimation declined from 11.0–11.9% (using the original Aqua–Terra data) to 1.0–1.5%. The method is able to improve the overall accuracy of the 8-day MODIS snow product from originally 78% to 89% and even to 93% during cloud free periods.
ISSN:0022-1694
1879-2707
DOI:10.1016/j.jhydrol.2015.12.065