NORMALIZED DIFFERENCE SNOW INDEX SIMULATION FOR SNOW-COVER MAPPING IN FOREST BY GEOSAIL MODEL

The snow-cover mapping in forest area is always one of the difficult points for optical satellite remote sensing. To investigate reflectance variability and to improve the mapping of snow in forest area, GeoSail model was used to simulate the reflectance of a snow-covered forest. Using this model, the effects of varying canopy density, solar illumination and view geometry on the performance of the MODIS (Moderate-resolution Imaging Spectroradiometer) snow-cover mapping algorithm were investigated. The relationship between NDSI (Normalized Difference Snow Index), NDVI (Normalized Difference Vegetation Index) and snow fraction was discussed in detail. Results indicated that the weak performance would be achieved if fixed criteria were used for different regions especially in the complicated land cover components. Finally, some suggestions to MODIS SNOWMAP algorithm were put forward to improve snow mapping precision in forest area based on the simulation, for example, new criteria should be used in coniferous forest, that is, NDSI greater than 0.3 and NDVI greater than zero. Otherwise, a threshold on view zenith angle may be used in the criteria such as 45°.

Estimating Ground Snow Load Based on Ground Snow Depth and Climatological Elements for Snow Hazard Assessment in Northeastern China

Extreme snow loads can collapse roofs.This load is calculated based on the ground snow load(that is,the snow water equivalent on the ground).However,snow water equivalent(SWE) measurements are unavailable for most sites,while the ground snow depth is frequently measured and recorded.A new simple practical algorithm was proposed in this study to evaluate the SWE by utilizing ground snow depth,precipitation data,wind speed,and air temperature.For the evaluation,the precipitation was clas sified as snowfall or rainfall according to the air temperature,the snowfall or rainfall was then corrected for measurement error that is mainly caused by wind-induced undercatch,and the effect of snow water loss was considered.The developed algorithm was applied and validated using data from57 meteorological stations located in the northeastern region of China.The annual maximum SWE obtained based on the proposed algorithm was compared with that obtained from the actual SWE measurements.The return period values of the annual maximum ground snow load were estimated and compared to those obtained according to the procedure suggested by the Chinese structural design code.The comparison indicated that the use of the proposed algorithm leads to a good estimated SWE or ground snow load.Its use allowed the estimation of the ground snow load for sites without SWE measurement and facilitated snow hazard mapping.