Fractional snow cover mapping through artificial neural network analysis of MODIS surface reflectance

Accurate areal measurements of snow cover extent are important for hydrological and climate modeling. The traditional method of mapping snow cover is binary where a pixel is considered either snow-covered or snow-free. Fractional snow cover (FSC) mapping can achieve a more precise estimate of areal snow cover extent by estimating the fraction of a pixel that is snow-covered. The most common snow fraction methods applied to Moderate Resolution Imaging Spectroradiometer (MODIS) images have been spectral unmixing and an empirical Normalized Difference Snow Index (NDSI). Machine learning is an alternative for estimating FSC as artificial neural networks (ANNs) have been successfully used for estimating the subpixel abundances of other surfaces. The advantages of ANNs are that they can easily incorporate auxiliary information such as land cover type and are capable of learning nonlinear relationships between surface reflectance and snow fraction. ANNs are especially applicable to mapping snow cover extent in forested areas where spatial mixing of surface components is nonlinear. This study developed a multilayer feed-forward ANN trained through backpropagation to estimate FSC using MODIS surface reflectance, NDSI, Normalized Difference Vegetation Index (NDVI) and land cover as inputs. The ANN was trained and validated with higher spatial-resolution FSC maps derived from Landsat Enhanced Thematic Mapper Plus (ETM+) binary snow cover maps. Testing of the network was accomplished over training and independent test areas. The developed network performed adequately with RMSE of 12% over training areas and slightly less accurately over the independent test scenes with RMSE of 14%. The developed ANN also compared favorably to the standard MODIS FSC product. The study also presents a comprehensive validation of the standard MODIS snow fraction product whose performance was found to be similar to that of the ANN.     

View angle effects on MODIS snow mapping in forests

Binary snow maps and fractional snow cover data are provided routinely from MODIS (Moderate Resolution Imaging Spectroradiometer). This paper investigates how the wide observation angles of MODIS influence the current snow mapping algorithm in forested areas. Theoretical modeling results indicate that large view zenith angles (VZA) can lead to underestimation of fractional snow cover (FSC) by reducing the amount of the ground surface that is viewable through forest canopies, and by increasing uncertainties during the gridding of MODIS data. At the end of the MODIS scan line, the total modeled error can be as much as 50% for FSC. Empirical analysis of MODIS/Terra snow products in four forest sites shows high fluctuation in FSC estimates on consecutive days. In addition, the normalized difference snow index (NDSI) values, which are the primary input to the MODIS snow mapping algorithms, decrease as VZA increases at the site level. At the pixel level, NDSI values have higher variances, and are correlated with the normalized difference vegetation index (NDVI) in snow covered forests. These findings are consistent with our modeled results, and imply that consideration of view angle effects could improve MODIS snow monitoring in forested areas.     

Subpixel monitoring of the seasonal snow cover with MODIS at 250 m spatial resolution in the Southern Alps of New Zealand: Methodology and accuracy assessment

This study describes a comprehensive method to produce routinely regional maps of seasonal snow cover in the Southern Alps of New Zealand (upper Waitaki basin) on a subpixel basis, and with the MODerate Resolution Imaging Spectroradiometer (MODIS). The method uses an image fusion algorithm to produce snow maps at an improved 250 m spatial resolution in addition to the 500 m resolution snow maps. An iterative approach is used to correct imagery for both atmospheric and topographic effects using daily observations of atmospheric parameters. The computation of ground spectral reflectance enabled the use of image-independent end-members in a constrained linear unmixing technique to achieve a robust estimation of subpixel snow fractions. The accuracy of the snow maps and performance of the algorithm were assessed carefully using eight pairs of synchronic MODIS/ASTER images. ‘Pixel-based’ metrics showed that subpixel snow fractions were retrieved with a Mean Absolute Error of 6.8% at 250 m spatial resolution and 5.1% after aggregation at 500 m spatial resolution. In addition, a ‘feature-based’ metric showed that 90% of the snowlines were depicted generally within 300 m and 200 m of their correct position for the 500-m and 250-m spatial resolution snow maps, respectively. A dataset of 679 maps of subpixel snow fraction was produced for the period from February 2000 to May 2007. These repeated observations of the seasonal snow cover will benefit the ongoing effort to model snowmelt runoff in the region and to improve the estimation and management of water resources.     

东北地区MODIS亚像元积雪覆盖率反演及验证

以中巴资源卫星数据作为地面“真值”影像,根据东北地区地理环境与气候特点对Salomoson亚像元积雪覆盖率模型参数进行修正,反演东北地区MODIS像元积雪覆盖率,并用不同方案对模型的稳定性和精度进行分析。研究结果表明,经修正后的Salomoson亚像元积雪覆盖率反演模型对不同地貌--景观单元具有稳定性,其中较小的波动源于积雪物理性质差异、大气效应、积雪影像分类误差及影像配准误差。在东北平原区,NDSI值在0.52~0.65时,模型反演精度高,但反演雪盖率总体偏低,主要是由NDSI基于对波段反射率的非线性转换引起的;雪盖率高估的像元主要分布在城区外围以及农村居民点,而覆盖城区、乡、镇以及居民点之间道路的像元雪盖率误差小,其原因是人类活动频率影响像元内积雪组分与非积雪组分的光谱特性的差异程度。与MODIS雪产品进行对比分析,积雪覆盖率提供较传统雪盖制图更加丰富的信息,然而对林区冠层下积雪覆盖二者均未给出准确估计。     

Enhanced algorithm for estimating snow depth from geostationary satellites

Observations in the visible and infrared spectral bands from the Imager instrument onboard Geostationary Operational Environmental Satellite (GOES) have been used to derive snow depth. The technique makes use of correlation between depth of the snow pack and satellite-derived subpixel fractional snow cover. Previous efforts to infer snow depth from satellite data with this technique were focused on grasslands and croplands, where the snow depth/snow fraction relationship is most pronounced. In this paper we improve the retrieval algorithm to extend snow depth estimates to forested areas. The enhanced algorithm accounts for the tree cover fraction and for the type of forest, deciduous or coniferous.The developed technique was used to derive maps of snow depth over mid-latitude areas of North America during winter seasons of 2003-2004 and 2004-2005. Satellite-based snow depth maps were produced daily at 4 km spatial resolution. To validate the retrievals we compared them with surface observations of snow depth and with the snow depth analysis prepared at the NOAA National Operational Hydrological Remote Sensing Center (NOHRSC). The estimated retrieval error was about 30% for snow depths below 30 cm and increased to 50% for snow depths ranging from 30 to 50 cm. Snow depth retrievals were limited to scenes with less than 80% deciduous forest cover fraction and less than 50% needle leaf forest cover.     

A new water-resistant snow index for the detection and mapping of snow cover on a global scale

An up-to-date spatio-temporal change analysis of global snow cover is essential for better understanding of climate–hydrological interactions. The normalized difference snow index (NDSI) is a widely used algorithm for the detection and estimation of snow cover. However, NDSI cannot discriminate between snow cover and water bodies without use of an external water mask. A stand-alone methodology for robust detection and mapping of global snow cover is presented by avoiding external dependency on the water mask. A new spectral index called water-resistant snow index (WSI) with the capability of exhibiting significant contrast between snow cover and other cover types, including water bodies, was developed. WSI uses the normalized difference between the value and hue obtained by transforming red, green, and blue, (RGB) colour composite images comprising red, green, and near-infrared bands into a hue, saturation, and value (HSV) colour model. The superiority of WSI over NDSI is confirmed by case studies conducted in major snow regions globally. Snow cover was mapped by considering monthly variation in snow cover and availability of satellite data at the global scale. A snow cover map for the year 2013 was produced at the global scale by applying the random walker algorithm in the WSI image supported by the reference data collected from permanent snow-covered and non-snow-covered areas. The resultant snow-cover map was compared to snow cover estimated by existing maps: MODIS Land Cover Type Product (MCD12Q1 v5.1, 2012), Global Land Cover by National Mapping Organizations (GLCNMO v2.0, 2008), and European Space Agency’s GlobCover 2009. A significant variation in snow cover as estimated by different maps was noted, and was was attributed to methodological differences rather than annual variation in snow cover. The resultant map was also validated with reference data, with 89.46% overall accuracy obtained. The WSI proposed in the research is expected to be suitable for seasonal and annual change analysis of global snow cover.     

Using MODIS snow cover maps in modeling snowmelt runoff process in the eastern part of Turkey

Water perhaps is the most valuable natural asset in the Middle East as it was a historical key for settlement and survival in Mesopotamia, “the land between two rivers”. At present, the Euphrates and Tigris are the two largest trans-boundary rivers in Western Asia where Turkey, Syria, Iran, Iraq and Saudi Arabia are the riparian countries. The Euphrates and Tigris basins are largely fed from snow precipitation whereby nearly two-thirds occur in winter and may remain in the form of snow for half of the year. The concentration of discharge mainly from snowmelt during spring and early summer months causes not only extensive flooding, inundating large areas, but also the loss of much needed water required for irrigation and power generation purposes during the summer season. Accordingly, modeling of snow-covered area in the mountainous regions of Eastern Turkey, as being one of the major headwaters of Euphrates-Tigris basin, has significant importance in order to forecast snowmelt discharge especially for energy production, flood control, irrigation and reservoir operation optimization.A pilot basin, located on the upper Euphrates River, is selected where five automated meteorological and snow stations are installed for real time operations. The daily snow cover maps obtained from Moderate Resolution Imaging Spectroradiometer MODIS at 500 m resolution are compared with ground information for the winter of 2002-2003 both during accumulation and ablation and at accumulation stage for the winter of 2003-2004. The snow presence on the ground is determined from the snow courses performed. Such measurements were made at 19 points in and around the upper Euphrates River in Turkey and at 20 points in the upper portion of the pilot basin for the winters of 2002-2003 and 2003-2004, respectively. Comparison of MODIS snow maps with in situ measurements over the snow season show good agreement with overall accuracies ranging between 62% and 82% considering the shift in the days of comparison. The main reasons to have disagreement between MODIS and in situ data are the high cloud cover frequency in the area and the current version of the MODIS cloud-mask that appears to frequently map edges of snow-covered areas and land surfaces. The effect of elevation and land cover types on validation of MODIS snow cover maps is also analyzed. In order to minimize the cloud cover and maximize the snow cover, MODIS-8 daily snow cover products are used in deriving the snow depletion curve, which is one of the input parameters of the snowmelt runoff model (SRM). The initial results of modeling process show that MODIS snow-covered area product can be used for simulation and also for forecasting of snowmelt runoff in basins of Turkey.     

Retrieval of subpixel snow covered area, grain size, and albedo from MODIS

We describe and validate a model that retrieves fractional snow-covered area and the grain size and albedo of that snow from surface reflectance data (product MOD09GA) acquired by NASA's Moderate Resolution Imaging Spectroradiometer (MODIS). The model analyzes the MODIS visible, near infrared, and shortwave infrared bands with multiple endmember spectral mixtures from a library of snow, vegetation, rock, and soil. We derive snow spectral endmembers of varying grain size from a radiative transfer model specific to a scene's illumination geometry; spectra for vegetation, rock, and soil were collected in the field and laboratory. We validate the model with fractional snow cover estimates from Landsat Thematic Mapper data, at 30 m resolution, for the Sierra Nevada, Rocky Mountains, high plains of Colorado, and Himalaya. Grain size measurements are validated with field measurements during the Cold Land Processes Experiment, and albedo retrievals are validated with in situ measurements in the San Juan Mountains of Colorado. The pixel-weighted average RMS error for snow-covered area across 31 scenes is 5%, ranging from 1% to 13%. The mean absolute error for grain size was 51 µm and the mean absolute error for albedo was 4.2%. Fractional snow cover errors are relatively insensitive to solar zenith angle. Because MODSCAG is a physically based algorithm that accounts for the spatial and temporal variation in surface reflectances of snow and other surfaces, it is capable of global snow cover mapping in its more computationally efficient, operational mode.     

Comparing MODIS and ETM+ data for regional and global land classification

Nearly simultaneous reflectance data sets from the Landsat 7 Enhanced Thematic Mapper Plus (ETM+), at 30-m resolution, and the Terra satellite instrument MODIS, at 500-m resolution, are compared for their ability to map fractional coverage of surface types over large areas. Lower spatial resolution MODIS classification results are generally comparable those of ETM+, with discrepancies for some regions with mixed surface types. Analysis of laboratory and field spectra suggests an ambiguity, the “brightness ambiguity”, which can prevent accurate area estimation of pixels having two or more surface types. This ambiguity, plus general mathematical inversion issues, can account for the discrepancy. Thus, occasional high-resolution measurements, as from Landsat 7, are necessary to refine estimations of large area surface types from MODIS and similar lower spatial resolution instruments.     

Monitoring snow cover variability in an agropastoral area in the Trans Himalayan region of Nepal using MODIS data with improved cloud removal methodology

Monitoring the extent and pattern of snow cover in the dry, high altitude, Trans Himalayan region (THR) is significant to understand the local and regional impact of ongoing climate change and variability. The freely available Moderate Resolution Imaging Spectroradiometer (MODIS) snow cover images, with 500 m spatial and daily temporal resolution, can provide a basis for regional snow cover mapping, monitoring and hydrological modelling. However, high cloud obscuration remains the main limitation. In this study, we propose a five successive step approach — combining data from the Terra and Aqua satellites; adjacent temporal deduction; spatial filtering based on orthogonal neighbouring pixels; spatial filtering based on a zonal snowline approach; and temporal filtering based on zonal snow cycle — to remove cloud obscuration from MODIS daily snow products. This study also examines the spatial and temporal variability of snow cover in the THR of Nepal in the last decade. Since no ground stations measuring snow data are available in the region, the performance of the proposed methodology is evaluated by comparing the original MODIS snow cover data with least cloud cover against cloud-generated MODIS snow cover data, filled by clouds of another densely cloud-covered product. The analysis indicates that the proposed five-step method is efficient in cloud reduction (with average accuracy of > 91%). The results show very high interannual and intra-seasonal variability of average snow cover, maximum snow extent and snow cover duration over the last decade. The peak snow period has been delayed by about 6.7 days per year and the main agropastoral production areas of the region were found to experience a significant decline in snow cover duration during the last decade.     

MODIS雪产品的计算及其与NSIDC雪产品的比较--以青藏高原雪覆盖为例

以青藏高原上的两个区域、选择不同时相的MODIS影像，运用NDSI、NDVI等指标，结合MODIS云腌膜数据，计算得到MODIS的雪产品；并将其与美国NSIDC的雪产品进行比较。结果表明，计算得到的MODIS雪产品中雪的覆盖区域与NSIDC雪产品中的雪覆盖区域基本一致，但计算得到的雪覆盖面积比对应的NSIDC雪产品中的雪覆盖面积略大3％。计算得到的雪产品与NSIDC雪产品在同样经纬度范围内图像位置有偏差。范围越小，位置偏差越明显。对图像进行经度偏移后，取共同区域进行像素差值比较，结果，计算得到的雪产品与NSIDC雪产品判断一致的像素约83．12％。     

A blended global snow product using visible,passive microwave and scatterometer satellite data

A joint US Air Force/National Aeronautics and Space Administration (NASA) blended global snow product that uses Earth Observation System Moderate Resolution Imaging Spectroradiometer (MODIS), Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E) and Quick Scatterometer (QuikSCAT or QSCAT) data has been developed. Existing snow products derived from these sensors have been blended into a single, global, daily, user-friendly product by using a newly developed Air Force Weather Agency (AFWA)/NASA Snow Algorithm (ANSA). This initial blended snow product uses minimal modelling to expeditiously yield improved snow products, which include, or will include, snow-cover extent, fractional snow cover, snow water equivalent (SWE), onset of snowmelt and identification of actively melting snow cover. The blended snow products are currently 25-km resolution. These products are validated with data from the lower Great Lakes region of the USA, from Colorado obtained during the Cold Land Processes Experiment (CLPX), and from Finland. The AMSR-E product is especially useful in detecting snow through clouds; however, passive microwave data miss snow in those regions where the snow cover is thin, along the margins of the continental snowline, and on the lee side of the Rocky Mountains, for instance. In these regions, the MODIS product can map shallow snow cover under cloud-free conditions. The confidence for mapping snow-cover extent is greater with the MODIS product than with the microwave product when cloud-free MODIS observations are available. Therefore, the MODIS product is used as the default for detecting snow cover. The passive microwave product is used as the default only in those areas where MODIS data are not applicable due to the presence of clouds and darkness. The AMSR-E snow product is used in association with the difference between ascending and descending satellite passes or diurnal-amplitude variations (DAV) to detect the onset of melt, and a QSCAT product will be used to map areas of snow that are actively melting.     

Field-based spectral reflectance measurements of seasonal snow cover in the Indian Himalaya

In the present study, spectroradiometer (350–2500 nm) experiments are carried out in the field to understand the influence of snow grain size, contamination, moisture, ageing, snow depth, slope / aspect on spectral reflectance and to determine the sensitive wavelengths for mapping of snow and estimation of snow characteristics using satellite data. The observations suggest that, due to ageing and grain-size variation, the maximum variations in reflectance are observed in the near-infrared region, i.e. around 1040–1050 nm. For varying contamination and snow depth, the maximum variations are observed in the visible region, i.e. around 470 and 590 nm, respectively. For the moisture changes, the maximum variations are observed around 980 and 1160 nm. Based on the spectral signatures of seasonal snow, the normalized difference snow index (NDSI) is studied, and snow indexes, such as grain and contamination indexes, are proposed. The study also suggests that the NDSI increases with ageing, grain size and moisture content. The NDSI values remain constant with variations in slope and aspect. Attempts are made to estimate seasonal snow characteristics using multispectral Advanced Wide Field Sensor (AWiFS) Indian Remote Sensing (IRS-P6) and Moderate Resolution Imaging Spectroradiometer (MODIS) Terra satellite data and validated with snow-meteorological observatory data of the study area.     

Toward improved daily snow cover mapping with advanced combination of MODIS and AMSR-E measurements

Taking three snow seasons from November 1 to March 31 of year 2002 to 2005 in northern Xinjiang, China as an example, this study develops a new daily snow cover product (500 m) through combining MODIS daily snow cover data and AMSR-E daily snow water equivalent (SWE) data. By taking advantage of both high spatial resolution of optical data and cloud transparency of passive microwave data, the new daily snow cover product greatly complements the deficiency of MODIS product when cloud cover is present especially for snow cover product on a daily basis and effectively improves daily snow detection accuracy. In our example, the daily snow agreement of the new product with the in situ measurements at 20 stations is 75.4%, which is much higher than the 33.7% of the MODIS daily product in all weather conditions, even a little higher than the 71% of the MODIS 8-day product (cloud cover of ~ 5%). Our results also indicate that i) AMSR-E daily SWE imagery generally agrees with MOD10A1 data in detecting snow cover, with overall agreement of 93.4% and snow agreement of 96.6% in the study area; ii) AMSR-E daily SWE imagery underestimates the snow covered area (SCA) due to its coarse spatial resolution; iii) The new snow cover product can better and effectively capture daily SCA dynamics during the snow seasons, which plays a significant role in reduction, mitigation, and prevention of snow-caused disasters in pastoral areas.     

Mapping global urban areas using MODIS 500-m data: New methods and datasets based on ‘urban ecoregions’

Although cities, towns and settlements cover only a tiny fraction (< 1%) of the world's surface, urban areas are the nexus of human activity with more than 50% of the population and 70-90% of economic activity. As such, material and energy consumption, air pollution, and expanding impervious surface are all concentrated in urban areas, with important environmental implications at local, regional and potentially global scales. New ways to measure and monitor the built environment over large areas are thus critical to answering a wide range of environmental research questions related to the role of urbanization in climate, biogeochemistry and hydrological cycles. This paper presents a new dataset depicting global urban land at 500-m spatial resolution based on MODIS data (available at http://sage.wisc.edu/urbanenvironment.html). The methodological approach exploits temporal and spectral information in one year of MODIS observations, classified using a global training database and an ensemble decision-tree classification algorithm. To overcome confusion between urban and built-up lands and other land cover types, a stratification based on climate, vegetation, and urban topology was developed that allowed region-specific processing. Using reference data from a sample of 140 cities stratified by region, population size, and level of economic development, results show a mean overall accuracy of 93% (k = 0.65) at the pixel level and a high level of agreement at the city scale (R² = 0.90).     

An application of MODIS data to snow cover monitoring in a pastoral area: A case study in Northern Xinjiang, China

Snow is an important land cover on the earth's surface. It is characterized by its changing nature. Monitoring snow cover extent plays a significant role in dynamic studies and prevention of snow-caused disasters in pastoral areas. Using NASA EOS Terra/MODIS snow cover products and in situ observation data during the four snow seasons from November 1 to March 31 of year 2001 to 2005 in northern Xinjiang area, the accuracy of MODIS snow cover mapping algorithm under varied snow depth and land cover types was analyzed. The overall accuracy of MODIS daily snow cover mapping algorithm in clear sky condition is high at 98.5%; snow agreement reaches 98.2%, and ranges from 77.8% to 100% over the 4-year period for individual sites. Snow depth (SD) is one of the major factors affecting the accuracy of MODIS snow cover maps. MODIS does not identify any snow for SD less than 0.5 cm. The overall accuracy increases with snow depth if SD is equal to or greater than 3 cm, and decreases for SD below 3 cm. Land cover has an important influence in the accuracy of MODIS snow cover maps. The use of MOD10A1 snow cover products is severely affected by cloud cover. The 8-day composite products of MOD10A2 can effectively minimize the effect of cloud cover in most cases. Cloud cover in excess of 10% occurs on 99% of the MOD10A1 products and 14.7% of the MOD10A2 products analyzed during the four snow seasons. User-defined multiple day composite images based on MOD10A1, with flexibilities of selecting composite period, starting and ending date and composite sequence of MOD10A1 products, have an advantage in effectively monitoring snow cover extent for regional snow-caused disasters in pastoral areas.     

Retrieval of snow grain size over Greenland from MODIS

This paper presents a new automatic algorithm to derive optical snow grain size at 1 km resolution using Moderate Resolution Imaging Spectroradiometer (MODIS) measurements. The retrieval is conceptually based on an analytical asymptotic radiative transfer model which predicts spectral bidirectional snow reflectance as a function of the grain size and ice absorption. The snow grains are modeled as fractal rather than spherical particles in order to account for their irregular shape. The analytical form of solution leads to an explicit and fast retrieval algorithm. The time series analysis of derived grain size shows a good sensitivity to snow melting and snow precipitation events. Pre-processing is performed by a Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm, which includes gridding MODIS data to 1 km resolution, water vapor retrieval, cloud masking and an atmospheric correction. MAIAC cloud mask is a new algorithm based on a time series of gridded MODIS measurements and an image-based rather than pixel-based processing. Extensive processing of MODIS TERRA data over Greenland shows a robust discrimination of clouds over bright snow and ice. Because in-situ grain size measurements over Greenland were not available at the time of this work, the validation was performed using data of Aoki et al. (Aoki, T., Hori, M., Motoyoshi, H., Tanikawa, T., Hachikubo, A., Sugiura, K., et al. (2007). ADEOS-II/GLI snow/ice products — Part II: Validation results using GLI and MODIS data. Remote Sensing of Environment, 111, 274-290) collected at Barrow (Alaska, USA), and Saroma, Abashiri and Nakashibetsu (Japan) in 2001-2005. The retrievals correlate well with measurements in the range of radii ~ 0.1-1 mm, although retrieved optical diameter may be about a factor of 1.5 lower than the physical measured diameter. As part of validation analysis for Greenland, the derived grain size from MODIS over selected sites in 2004 was compared to the microwave brightness temperature measurements of SSM/I radiometer which is sensitive to the amount of liquid water in the snowpack. The comparison showed a good qualitative agreement, with both datasets detecting two main periods of snowmelt. Additionally, MODIS grain size was compared with predictions of the snow model CROCUS driven by measurements of the automatic weather stations of the Greenland Climate Network. We found that the MODIS value is on average a factor of two smaller than CROCUS grain size. This result agrees with the direct validation analysis indicating that the snow reflectance model may need a “calibration” factor of ~ 1.5 for the retrieved grain size to match the physical snow grain size. Overall, the agreement between CROCUS and MODIS results was satisfactory, in particular before and during the first melting period in mid-June. Following detailed time series analysis of snow grain size for four permanent sites, the paper presents maps of this important parameter over the Greenland ice sheet for the March-September period of 2004.     

Evaluation of MODIS snow cover and cloud mask and its application in Northern Xinjiang, China

Using five-year (2001-2005) ground-observed snow depth and cloud cover data at 20 climatic stations in Northern Xinjiang, China, this study: 1) evaluates the accuracy of the 8-day snow cover product (MOD10A2) from the Moderate Resolution Imaging Spectroradiometer (MODIS)/Terra satellite, 2) generates a new snow cover time series by separating the MODIS cloud masked pixels as snow and land, and 3) examines the temporal variability of snow area extent (SAE) and correlations of air temperature and elevation with SAE. Results show that, under clear sky conditions, the MOD10A2 has high accuracies when mapping snow (94%) and land (99%) at snow depth ≥ 4 cm, but a very low accuracy (< 39%) for patchy snow or thin snow depth (< 4 cm). Most of the patchy snow is misclassified as land. The mean accuracy of the cloud mask used in MOD10A2 for December, January and February is very low (19%). Based on the ratio of snow to land of ground observations in each month, the new snow cover time series generated in this study provides a better representation of actual snow cover for the study area. The SAE (%) time series exhibits similar patterns during six hydrologic years (2001-2006), even though the accumulation and melt periods do not exactly coincide. The variation of SAE is negatively associated with air temperature over the range of − 10 °C to 5 °C. An increase in elevation generally results in longer periods of snow cover, but the influence of elevation on SAE decreases as elevation exceeds 4 km in the Ili River Watershed (IRW). The number of days with snow cover shows either a decreasing trend or no trend in the IRW and the entire study area in the study period. This result is inconsistent with a reported increasing trend based on limited in situ observations. Long-term continuance of the MODIS snow cover product is critical to resolve this dilemma because the in situ observations appear to undersample the region.     

Retrieval of sub-pixel snow cover information in the Himalayan region using medium and coarse resolution remote sensing data

Numerous constrained and unconstrained algorithms have been used to retrieve sub-pixel snow-cover information quantitatively using medium and coarse spatial resolution multispectral images from the Advanced Wide Field Sensor (AWiFS) and Moderate Resolution Imaging Spectrometer (MODIS) sensors over the Himalayan region. Both the methods give slow convergence rates and inaccurate estimation of sub-pixel components analysed using root mean square (RMS) error and image deviation. Multiplicative iterative algorithms such as the Expectation Maximization Maximum Likelihood Method (EMML) and the Image Space Reconstruction Algorithm (ISRA) based on the minimization of least squares and Kullback–Leibler distances have been attempted to compute the endmembers' abundances in unmixing of satellite data. In this paper we discuss the eigenvalues of minimum noise fraction (MNF) transformation bands, data noise removal using MNF transformation and selection of pure endmembers using satellite images. The normalized difference snow index (NDSI) is also estimated using field spectral reflectance results and satellite images in green and shortwave infrared (SWIR) wavelength regions in order to carry out a comparative analysis for its variations with sub-pixel snow cover fractions. The present analysis shows the advantage of iterative over direct (constrained and unconstrained) methods; constraints are easily handled and allow better regularization of the solution for the ill-conditioned cases. Iterative methods are found to be faster compared to those of direct methods and can be used operationally for all resolution data for accurate estimation of sub-pixel snow cover.     

基于MODIS数据的雪情分析研究

利用基于MODIS数据的NDSI和基础地理信息数据对西藏地区进行雪情分析，结果表明这是可行的。MODIS数据在空间分辨率和波谱分辨率上优于AVHRR，利用MODIS数据进行雪情分析和其他监测会更有利。本研究说明利用MODIS数据可以进行大面积雪情分析、雪灾监测。