Estimation of fractional snow cover based on FY-2E/VISSR over … · Estimation of fractional snow cover based on FY-2E/VISSR over the Tibetan Plateau Lingmei Jiang 1, Jiancheng Shi
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Estimation of fractional snow cover based on FY-2E/VISSR over the Tibetan Plateau
Lingmei Jiang 1, Jiancheng Shi 2, Juntao Yang 1
1. State Key Laboratory of Remote Sensing Science, School of Geography, Beijing Normal University, Beijing 100875,
The remained 10 ETM+ fractional snow cover is used to validate fractional snow cover algorithm developed in this
work. Table 3 indicated that the models based on DSI show better performance than others based on RSI and NDSI. This
is consistent with the developed fitting models in Table 2. Average R-square of the DSI-based models is approximately
0.61, with RMSE approximately 0.17. R-square of two RSI-based models is approximately 0.39 and 0.36 for
NDSI-based model. RMSE of RSI-based model and NDSI-based models is approximately 0.21. In the 6 fitting models,
Model 1 seems to be the most suitable model for estimating fractional snow cover based on VISSR DSI, which shows the
highest R-square and lowest RMSE.
Figure 3 presents VISSR fractional snow cover estimated by 6 snow indices models and ETM+ fractional snow
cover (on April 26th, 2011), ETM+ RGB composite image with band 7, 4 and 2. This comparison of VISSR FSC of 6
models with ETM+ FSC also demonstrated that models based on DSI present better performance in estimating fractional
snow cover. When compared with ETM+ FSC, the three DSI-based models show little under-estimation of FSC in some
regions. However, FSC images of the models based on RSI and NDSI show over-estimation of FSC overall. In general,
the VISSR fractional snow cover estimation results validated by 10 ETM+ scenes indicate that the models based on the
DSI perform better than the other models.
The accuracy of FY-2E/VISSR snow cover image is 88.67% when compared with the meteorological station snow
depth observations in 2010 and 2011 winter seasons. The FY-2E/VISSR snow cover image presented slight
over-estimation of snow cover over the Tibetan Plateau. And the accuracy was similar to IMS snow cover products,
which is 85.76% [6]. The overall accuracy of FY-2E/VISSR snow cover image is lower than MODIS snow cover image,
due to its coarse spatial resolution and spectral bands. In addition, mixed pixel problem should be considered in FY2
snow cover mapping
Figure 3. Fractional snow cover image estimated by
VISSR of 6 models and ETM+ fractional snow cover
image, ETM+ RGB image using band 7, 4 and 2
(Landsat-7 scene on April 26th, 2011)
Table 3. Comparisons of VISSR FSC with ETM+ FSC of
6 models
ModelSnow Index
used R-square RMSE
Regression of VISSR FSC (x) with ETM+
FSC (y)
1 DSI 0.6196 0.1705 y=1.0511*x-0.0298
2 DSI 0.6101 0.1725 y=1.1155*x-0.0778
3 DSI 0.6083 0.1763 y=1.0642*x-0.0184
4 RSI 0.3959 0.2168 y=0.8968*x+0.0812
5 RSI 0.3941 0.2173 y=0.9202*x+0.0738
6 NDSI 0.3586 0.2184 y=0.9482*x+0.0485
5. Conclusion
In this paper, we developed an empirical algorithm based on snow indices to estimate fractional snow cover from
FY-2E/VISSR. Because VISSR lacks of 1.6μm channel, snow indices including NDSI, RSI and DSI are estimated based
on 0.6μm and 3.9μm channels. The results show that the fitting models based on DSI with exponential and logarithmic
curves can be utilized to estimate FSC with high R-squares and low RMSE. When validated by ETM+ FSC, the fitting
model based on VISSR DSI with exponential curve fitting show the best performance, which R-square is about 0.62 and
RMSE is about 0.17. This work demonstrated the potential of snow fraction estimation from FY-2E/VISSR.
Acknowledgement
This study is partly supported by the National Basic Research Program of China (No. 2013CB733406) and National
Natural Science Foundation of China (No. 41171260). The authors would like to thank China Meteorological
Administration (CMA) and U.S. Geological Survey (USGS) for providing satellites data.
References 1. Metsämäki, S., Mattila, O., Pulliainen, J., Niemi, K., Luojus, K., and Böttcher, K. ”An optical reflectance model-based
method for fractional snow cover mapping applicable to continental scale”. Remote Sensing of Environment, 2012, 123, pp. 508-521.
2. Shreve, C.M., Okin, G.S., and Painter, T.H. “Indices for estimating fractional snow cover in the western Tibetan Plateau”. Journal of Glaciology, 2009,55, pp.737-745.
3. Lei, L., Zeng, Z., and Zhang, B. “Method for detecting snow lines from MODIS data and assessment of changes in the Nianqingtanglha Mountains of the Tibet Plateau”. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2012, 5(3), pp.769-776.
4. Zhu, J., Shi, J., and Wang, Y. “Subpixel snow mapping of the Qinghai–Tibet Plateau using MODIS data”. International Journal of Applied Earth Observation and Geoinformation, 2012, 18, pp.251-262.
5. Zhang, G., Xie, H., Yao, T., Liang, T., and Kang, S. “Snow cover dynamics of four lake basins over Tibetan Plateau using time series MODIS data (2001-2010)”. Water Resources Research, 2012, 48, W10529.
6. Yang, J., Jiang, L., Shi, J., Wu, F., Wang, S., and Kou, X.. “Evaluation and comparison of FY-2E VISSR, MODIS and IMS snow cover over the Tibetan Plateau”. Proceedings of Geoscience and Remote Sensing Symposium, IGARSS 2013, IEEE 2013 International. 21-26 July 2013, Melbourne, Australia , pp. 1182-1185.
7. Romanov, P., Tarpley, D., Gutman, G., and Carroll, T. “Mapping and monitoring of the snow cover fraction over North America”. Journal of geophysical research, 2003, 108, 14.
8. Lin, J., Feng, X., Xiao, P., Li, H., Wang, J., and Li, Y. “Comparison of snow indexes in estimating snow cover fraction in a mountainous area in Northwestern China”. IEEE Geoscience and Remote Sensing Letters, 2012, 9(4), pp.725-729.
9. Masek, J.G., Vermote, E.F., Saleous, N.E., Wolfe, R., Hall, F.G., Huemmrich, K.F., Gao, F., Kutler, J., and Lim, T. “A Landsat surface reflectance dataset for North America, 1990–2000”. IEEE Geoscience and Remote Sensing Letters, 2006,3(1), pp.68-72.