An Improved Snow Depth Retrieval Algorithm for AMSR2 Passive Microwave Data Based on Snow Survey Data in Northeast China

Snow plays an important role in global hydrological cycle and climate system. In addition, snow as a key variable factor is important indicator to climate change. In recent years, applying passive microwave remote sensing to retrieval snow depth has made greatly progress. Snow depth retrieval algorithms and snow products can provide spatial and temporal information on snow cover distribution, which is an important data source to snow monitor. Based on the survey project of snow cover characteristics and snow distribution in Northeast China, Advanced Microwave Scanning Radiometer 2 (AMSR2) brightness temperature data and MODIS product data are used in this paper. Original resolution brightness temperature observations are used in AMSR2 algorithm, except for the 89GHz channel which is resampled according to 36GHz footprint, and brightness temperature corrections made on the original measurements rather than using the aggregated brightness temperature observations. The original algorithm takes advantage of the expanded range of channels available on the AMSR2 instrument compared with the SSM/I and SMMR, and it undertakes a forest correction. However, the original AMSR2 snow depth retrieval algorithm greatly overestimate measured snow depth, especially in forest areas where is a challenge to retrieval snow depth, the error between 20 to 30cm. In order to improve the retrieval accuracy of snow depth in Northeast China, an improved snow depth retrieval algorithm in forest areas is proposed in this paper, the improved snow depth retrieval algorithm in forest areas is SD
f
= 1/ln(pol
36
) × (Tb
18V
− Tb
36V
)/(1 − fd × 0.6). The results proved that the error is less than 5cm after using the improved algorithm, the root mean square error and deviation are 4.82cm, 3.55cm, respectively.