Remote Sensing of Environment 114 (2010) 1662–1675

Contents lists available at ScienceDirect

Remote Sensing of Environment
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / r s e

Integrated assessment on multi-temporal and multi-sensor combinations for reducing cloud obscuration of MODIS snow cover products of the Pacific Northwest USA
Yang Gao a,b,c, Hongjie Xie b,⁎, Tandong Yao c, Chongsheng Xue a
a b c

Faculty of Earth Sciences, China University of Geosciences, Wuhan 430073, China Laboratory for Remote Sensing and Geoinformatics, University of Texas at San Antonio, San Antonio, TX 78249, USA Key Laboratory of Tibetan Environment Changes and Land Surface Processes, Institute of Tibetan Plateau Research, Chinese Academy of Sciences, Beijing 100085, China

a r t i c l e

i n f o

a b s t r a c t
MODIS (Moderate Resolution Imaging Spectroradiometer) snow cover products, of daily, freely available, worldwide spatial extent at medium spatial resolution, have been widely applied in regional snow cover and modeling studies, although high cloud obscuration remains a concern in some applications. In this study, various approaches including daily combination, adjacent temporal deduction, fixed-day combination, flexible multi-day combination, and multi-sensor combination are assessed to remove cloud obscuration while still maintain the temporal and spatial resolutions. The performance of the resultant snow cover maps are quantitatively evaluated against in situ observations at 244 SNOTEL stations over the Pacific Northwest USA during the period of 2006–2008 hydrological years. Results indicate that daily Terra and Aqua MODIS combination and adjacent temporal deduction can reduce cloud obscuration and classification errors although an annual mean of 37% cloud coverage remains. Classification errors in snow-covered months are actually small and tend to underestimate the snow cover. Primary errors of MODIS daily, fixed and flexible multi-day combination products occur during transient months. Flexible multi-day combination is an efficient approach to maintain the balance between temporal resolution and realistic estimation of snow cover extent since it uses two thresholds to control the combination processes. Multi-sensor combinations (daily or multi-day), taking advantage of MODIS high spatial resolution and AMSR-E cloud penetration ability, provide cloud-free products but bring larger image underestimation errors as compared with their MODIS counterparts. © 2010 Elsevier Inc. All rights reserved.

Article history: Received 8 December 2009 Received in revised form 18 February 2010 Accepted 21 February 2010 Keywords: Adjacent temporal deduction Fixed-day combination Flexible multi-day combination Multi-sensor combination Accuracy assessment

1. Introduction and background Snow cover is one of the most important climatic factors affecting surface albedo, energy balance and hydrological circulation. Snow cover mapping has been utilized in operational snowmelt, runoff forecasting, data assimilation and the calibration or validation of various hydrological models (Andreadis & Lettenmaier, 2006; Blöschl et al., 1991; Grayson et al., 2002; Parajka & Blöschl, 2008; Udnaes et al., 2007). Daily and freely available, worldwide spatial extent, and relatively high spatial resolution MODIS snow products have been widely applied in regional snow cover and modeling studies (Brown et al., 2007; Rodell & Houser, 2004). Many studies (i.e., Hall & Riggs, 2007; Liang et al., 2008a; etc.) have demonstrated that MODIS daily snow cover products in clear-sky condition are in quite good agreement with ground based observations or other satellite-based snow cover products, however high cloud obscuration is a real problem in using

⁎ Corresponding author. Tel.: + 1 210 458 5445; fax: + 1 210 458 4469. E-mail address: Hongjie.xie@utsa.edu (H. Xie). 0034-4257/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.rse.2010.02.017

MODIS snow cover products for various applications (Ault et al., 2006; Bitner et al., 2002; Klein & Barnett, 2003; Lopez et al., 2008; Simic et al., 2004; Tekeli et al., 2005; Wang et al., 2008; Xie et al., 2009; Zhou et al., 2005). Obtaining cloud-free or even cloud percentage less than 10% of daily MODIS images remains a challenge. Various approaches have been proposed to reduce cloud obscuration by altering the cloud mask, separating cloud-masked pixels, and applying spatial–temporal or multi-sensor combinations. For example, an improved cloud mask was used in the MODIS snow cover mapping algorithm to better separate snow from cloud (Riggs & Hall, 2003). The improvement, however, is usually small as the cloud indeed exists (Parajka & Blöschl, 2008). Wang et al. (2008) attempted to separate the cloud-masked pixels into “snow” and “land” based on the ratio of snow to land from ground observations in each month, which can estimate overall snow percentage but cannot identify actual snow distribution under cloud. Spatio-temporal or multi-sensor combinations are effective approaches to reduce cloud obscuration and identify snow distribution under cloud. Spatial approaches aim at replacing cloud pixels by the majority of non-cloud pixels in an eight pixel neighborhood (Parajka & Blöschl,

and forest in the mountainous areas. not for massive area of cloud cover (i. / Remote Sensing of Environment 114 (2010) 1662–1675 1663 2008. Finally. 2009). snow coverage and overall classification accuracy. Second. Multi-sensor approaches can generate cloud-free snow cover maps through the merging of MODIS and AMSR-E snow products at the expense of spatial resolution. First. this method may result in a composite product of over 8 days or even several weeks.2°W and from 41..5°N to 49. 2008a. 1). One disadvantage of this method is that it does not control the cloud percentage for the entire image area. Based on a predefined maximum cloud coverage threshold such as 10%.. 2009. Liang et al. These products have higher temporal resolution (average 2–3 days per image) and relatively low cloud coverage..e.. This method sacrifices spatial resolution due to the coarse spatial resolution of the AMSR-E snow product (25 km). 2008. 1.2. The mean annual precipitation is less than 200 mm in the northwest and almost 2800 mm in the central and eastern parts of this area. Gao et al. Rapid snowmelt events even cause major floods in this area (Kunkel et al. and to assess the performance of different products. Liang et al. some user-defined multi-day snow cover products are produced through combining MOD10A1 products in fixed temporal windows (Houborg et al...000 km2 and extends from 103°W to 124. It includes most of Washington. The central and eastern parts of this area are mountainous and are part of the Rocky Mountains. Xie et al..1.e.. Gao et al. Parajka & Blöschl.. scrub and grassland in the medium elevation regions. In situ observation datasets The in situ observation datasets (SNOTEL) used in this study include daily snow depth measurements at 244 stations from September 1. 1 shows the locations of these SNOTEL stations with mean annual snow cover duration for each station. (2010) develop a new method to fill the cloud gap based on the most recent cloud-free observations for each cloud pixel. Using the same algorithm. Tong et al. 2007). Temporal approaches merge daily or multi-day MODIS snow cover products to minimize cloud coverage and maximize snow coverage. These approaches largely reduce cloud coverage. when weather conditions remain overcast for over a week. To compare with satellite image observations. Fig. and iii) to analyze the tradeoff between image classification accuracy and cloud reduction. Yang et al. there are 244 SNOTEL stations covering a wide range of elevation zones. 2008b). Observations from 244 SNOwpack TELemetry (SNOTEL) stations are used to validate these products. thus it can provide opportunities for comparison and evaluation over various environmental conditions. ii) to evaluate and compare those improved snow cover products in terms of spatial extent of cloud coverage. Yang et al. calculated based on three hydrological Fig. but the fixed temporal windows reduce the products' ability to monitor snowfall events to some degrees. The area is selected for this study for several reasons. This can effectively reduce cloud blockage of daily MODIS snow products by ∼ 7%. other multi-day composite products with flexible starting and ending dates are produced (Wang et al. 2. Xie et al. All works performed in this paper are based on MODIS and AMSR-E data for the Pacific Northwest USA during three hydrological years from September 2005 to August 2008. Study area and data 2. or snow-free (no snow) otherwise. 2009. all the neighboring pixels are cloud-covered). . 2008b. The color symbols of stations represent the mean annual snow cover duration (days) in 2006– 2008 hydrological years. 2007a. Hall et al. 2009. 2005 to August 31.. the prevalent snowpack is an important water resource of this area.Y.5 mm). between temporal resolution and spatial resolution reservation. The total study area is ∼ 565. which provide comprehensive ground observation data. 2. which have a 3 hour difference. Elevation ranges from 0 to 4394 m. which is much higher than the 30. southwestern Montana.5°N. 2007.. Land use is mainly cultivated crops and pasture in the lowlands. Hall et al. The elevation of the western coast region is relatively low.1 in (or 2. and northeastern Wyoming (Fig. The snow classification accuracy of the MODIS and AMSR-E blended snow cover products is 85. northern and eastern Idaho. whose performances are suitable for different applications. 2006). northeastern Oregon. but ignore some special situations.6%. It combines 2 to 8 days of Terra/Aqua MODIS daily snow cover products (MOD10A1/MYD10A1) over defined periods... Topography of the study area and spatial distribution of 244 SNOTEL stations. The MODIS 8-day Snow Cover Product (MOD10A2/MYD10A2) is the representative of fixed-day combined products. 2007. Sorman et al.. For example. The mountainous areas are covered with snow from November to early May which highlights the importance of snow as a major water resource for the region. 2009. The objectives of this paper are i) to improve some current cloud reducing approaches and generate new daily or multi-day blended snow cover maps based on Terra/Aqua MODIS and Aqua AMSR-E snow products. 2006). can reduce ∼ 10–20% cloud obscuration (Parajka & Blöschl. Foster et al. by the sacrifice of temporal resolution. The purpose is to obtain various cloud-free snow cover products with reasonable temporal and spatial resolutions. Wang et al. This gap-filling strategy is a useful and dynamic method which uses all the nearest non-cloud observations.. 2008. except for volcanic areas. 2008. 2010).7% of MODIS daily products in all-sky condition (Liang et al. this area encompasses diverse topographic and climate regions. Daily combination of Terra and Aqua MODIS. Zhao & Fernandes.... Study area The study area is located in the Pacific Northwest USA. 2007. This approach takes advantage of both high spatial resolution of optical sensors and cloud penetration of passive microwave sensors (i. the in situ records are defined as snow cover (snow) when the value of snow depth is larger than or equal to 0. The method is really effective for scattered cloud cover. 2009).

the spatial distribution of these SNOTEL stations is expected to capture the variability of snowfall for this region. 15. 31. 2008 MOD10A1 total no. it's found that the total of MODIS-recorded snow-covered lake ice. Gao et al. 2009). 2. These images are reprojected into Albers equalarea conical projection using the HDF EOS-to-GeoTiff (HEG) tool (Taaheri et al. Fig. and missing data. 2008. follows the priority principle (Xie et al. A low-priority integer value is always replaced by a high-priority integer value. 2006 Dec. 2007 Apr. 3 shows the workflow chart of data processing. The composite rule. 2007). assuming that a pixel belongs to the cloud class in the image of the current day. in order to match with MODIS data. and vice versa. SWE value 0 represents snow-free land and is transformed to the no snow (land) category.3. 13. The AE_DySno classes are reclassified to three categories: snow. 18. Kawanishi et al. Li et al.4%. MODIS daily combination and adjacent temporal deduction Daily combination means that MOD10A1 (Terra) and MYD10A1 (Aqua) acquired on the same day are merged into one MODIS Daily Combined (MODISDC) snow cover product. 27. 2007).5% of the study area is above 3005 m elevation where the highest station is located. 1. multi-temporal (fixed-day and flexible multi-day). 3. 2005–Aug. Methods years' data from September. this study (Table 1).. It differs from direct replacement using previous snow mapping results with the loss of 1 day temporal resolution (Zhao & Fernandes. 2007 Dec. 2007 363 364 . 25% of the study area has an elevation lower than 1000 m while 8 SNOTEL stations are located in this elevation zone. Remote sensing data The snow cover images used in this study consist of MODIS daily snow cover products MOD10A1 and MYD10A1 (MODIS Terra/Aqua Snow Cover Daily L3 Global 500 m SIN GRID V005) (Hall et al. with the priority order being snow N no snow (land) N clouds. 16. 23. nearest day or over multi-day is used to reduce cloud obscuration. Hydrological year Sep.. covering a period of 3 years. three different situations may occur. 2007 Nov. no snow (land). and fill) do not occur. and multi-sensor combinations are tested to map snow cover while reducing cloud obscuration. 2. 17. and the data missing class is retained as data missing class. data missing and no decision data classes are merged into the cloud category. 2005 to August. 2008. 2009) and also from the spatial filtering which replaces cloud pixels with the majority class of the neighboring eight pixels in the same image (Parajka & Blöschl. 22. 362 Various approaches including daily combination. 365 AE_DySno missing date Nov. 2006 Aug. (2002). First. 2008). 31. ocean. 3. Only 1. The other remaining original classes do not occur in the SWE products of the study area during the period of study. 2007) from September 1. Details about the MODIS and AMSR-E snow products can be found in Lobl et al. 2005 AE_DySno total no. Fig... 253 (ice sheet). are used in this study (Table 1). no snow (land). Two MODIS tiles (H09V04 and H10V04) are needed to cover the entire study area. 2008 period. The snowcovered lake ice and snow classes are merged into a single snow category. the MY10A1 missing date / MYD10A1 total NO. data missing and no decision classes were less than 0. 248 (off-earth). The pixel values of AMSR-E SWE include 0–240 (multiplying by 2 for true SWE values in mm). 2004). 2008). 2007 365 365 Nov. 2005 Aug. 2007 Sep. and the cloud. / Remote Sensing of Environment 114 (2010) 1662–1675 Fig. 2006 Nov. For example. We first mosaic two tiles and then reproject them into Albers equal-area conical projection using the MODIS Reprojection Tool (MRT. 364 Sep. (2003). 2006–Aug. 2009. Then AMSR-E products with lower spatial resolution and higher retrieval rate are introduced. information from different MODIS products acquired on the same day. without any sacrifice of spatial or temporal resolution (similar as Gafurov & B'ardossy. 2005 to August 31. 2. and cloud.1664 Y. Hall and Riggs (2007) and Kelly et al. 1. the cloud pixel of the current day image is deduced as snow. the cloud pixel in the current day image is deduced as land. If only MOD10A1 is present (no MYD10A1 available on that day). Totals of 1090 MOD10A1 images and 1095 MYD10A1 images. 252 (land or snow impossible)...1. Adjacent temporal deduction is an approach to deduce the surface conditions under the remaining cloud pixels using the information of the preceding and following days on the same pixel basis. and (3) if the pixel values in the preceding and following day images are different (one snow. 31. (2003). 1. 254 (water) and 255 (data missing) (Kelly et al. (2) if the corresponding pixel values in both images of the preceding and following days are land. the land and inland water classes are merged into the no snow category. 2005–August. 2006 / Nov. (1) If the corresponding pixel values in both images of the preceding and following days are snow. 22. A total of 1091 AE_DySno images are used in Table 1 Remote sensing images used during September. no combination is performed and the MOD10A1 is directly used for the next process. 2 shows the percentage of different elevation zones of the entire study area and the frequency distribution of stations in different elevation zones. The pixel classes in those snow cover images are reclassified to three categories: snow. 23. The majority of the SNOTEL stations are distributed at elevations between 1000 m and 3000 m where most snowfalls occur. SWE values 1–240 representing snow-covered area are merged into the snow category. The elevations of the stations vary from 308 to 3005 m. The AMSR-E onboard Aqua satellite provides daily SWE products at 25 km spatial resolution. 2008 365 363 / Dec. In this study area. 2007–Aug. 28. similar to the generation of standard MOD10A2 snow cover products. detector saturated. 2006 MOD10A1 missing date Sep. adjacent temporal deduction. In adjacent temporal deduction. while the other four original classes (night. The percentage of different elevation zones to the entire study area and the station distribution in different elevation zones. Therefore. 2007b) and AMSR-E daily Snow Water Equivalent (SWE) product AE_DySno2 (AMSR-E/Aqua daily L3 Global Snow Water Equivalent EASE-Grids) (Kelly et al.

or the number of composite days equals to the threshold N. 2008b. When the station-pixel observation matches the station observation (i. and MODISMC8). if a station-pixel shows snow class.. / Remote Sensing of Environment 114 (2010) 1662–1675 1665 were named as MODIS Multi-day Combined (MODISMC) snow cover map. in these two periods. namely. Thus. are tested to replace cloud pixels by the most recent clear-sky observations of the same pixel.. respectively. the pixel values within the image gaps are replaced by the previous day's AMSR-E image (Liang et al. the cloud pixel of current day image remains as cloud. the adjacent temporal deduction method may lose its effectiveness towards cloud reduction. two types of accuracies are used: . we compared the composite image with observations from stations for individual days. we say the observation (or classification) of the image is in agreement. both expressed as below (in percentage): IU = IO = b a+b+c+d c a+b+c+d ð3Þ ð4Þ Fig. MODIS4DC. Parajka & Blöschl. Both are user-defined parameters that can be assigned depending on the application. three cloud-free multisensor snow cover products. The input image of this process is MODISDC and the resultant image is defined as MODIS adjacent temporal deduction (MODISATD) snow cover maps. and MODISMC8_AE. MODISMC4.2. 2000). 4.e. Generally the accuracy of MODIS only considers clear-sky (or cloud-free) condition. However. it is not in agreement. The resultant images from this process Ps = ð2Þ where Nc is the number of cloud pixels.. To quantitatively assess the classification accuracy of standard and combined images. The resultant images for fixed-day combination are named as MODIS2DC. image misclassification of snow as land) and image overestimation error (IO. and 8-day. or station-pixel no snow with and station no snow). the snow cover might change daily. in situ snow measurements are used as “truth value” for the image pixel (named as station-pixel) that includes such an in situ station. Nt is the total number of the pixels in the entire study area. Romanov et al. Image underestimation error (IU. 3. other land or cloud) or are both cloud. are therefore generated. The approach first calculates the cloud percentage (P) of the study area in the first input image. MODISMC4 for N as 4 days and MODISMC8 for N as 8 days. 2006). and the cloud percentage of the study area in the combined image is updated. A workflow chart of data processing. b.3. daily MODISATD_AE. Multi-temporal combination Two types of multi-temporal combination. 6-. Evaluation of the resultant snow cover maps The evaluation includes three parts: the spatial extent of cloud and snow. in order to compare with AMSR-E and MODIS/AMSR-E blended snow cover maps which have no cloud. c. It takes advantage of high spatial resolution and high accuracy in clear-sky condition of MODIS images and the cloud transparency ability of AMSR-E snow products (Liang et al. and MODIS8DC respectively for the 2-. For the snow accumulating and melting periods. 3. multi-day MODISMC4_AE. If it is less than the threshold P. 4-. 2010). d.. This assumption should hold true for a continuous snow-covered or snow-free period. The other is based on flexible temporal windows. They are expressed as below (in percentage): Pc = Nc Nt Ns Nt ð1Þ Fig. The agreement of remote sensing image classification is evaluated by overall accuracy and snow accuracy (Klein & Barnett. where a. Pu et al. For multi-day combined products. 2007.. Flexible multi-day combination approach is presented in Fig. The process will not stop until either the cloud percentage is less than the threshold P. The daily MODISATD snow cover maps are the inputs for these multi-temporal combinations. else. we say that image agrees in 3 days and disagrees in 1 day. A workflow chart of flexible multi-day combination algorithm. 2007). image misclassification of land as snow) are widely used in assessing the misclassification of remote sensing images (Parajka & Blöschl. 2008b. Gao et al. in the 4-day composite product. 2008b. it is combined with the second input image.4. we defined the P as 10% and N as 4 days or 8 days. In this study. 2003. the image directly outputs. station-pixel snow with station snow. while the corresponding station (SNOTEL) shows snow for 3 days and land for 1 day. The flexible temporal windows are controlled by two thresholds. Multi-sensor combination Multi-sensor combination uses AMSR-E to map those areas where MODIS cannot map because of clouds or polar nights (Foster et al. (3)–(10) represent the number of station-pixels in each particular classification category (Table 2). Gao et al. 2008.Y.. Based on the three MODIS snow cover products (MODISATD. fixed-day and flexible multi-day combinations. 6-. Since there are AMSR-E image gaps and the locations of these gaps change daily. the misclassification and classification accuracies against in situ observations.. Liang et al. Ns is the number of snow pixels. 4-. The first one is based on fixed temporal windows of 2-. The percentages of cloud (Pc) and snow (Ps) coverage over the entire study area are used to reflect the performance on cloud reduction and snow retrieval from different combinations. and 8-day combinations. P and N are userdefined thresholds and are predefined as 10% and 4 days or 8 days respectively in this study. 2008). e and f in Eqs. MODIS6DC. maximum cloud percentage (P) and maximum composite days (N). 3. For example. otherwise. 3. 4.

to the Terra/Aqua combined products (MODISDC). 5. which means that all four products tend to underestimate the snow extent. The snow accuracy in clear-sky (Sc) or in all-sky (Sa) conditions is respectively calculated by the number of station-pixels with snow that agreed with station snow divided by the total snow-covered station-pixels in clear-sky condition or in all-sky condition. the snow melting period). . to daily combined products. during 2006–2008 hydrological years. Aqua MODIS (MYD10A1). Mean percentage (monthly) of cloud coverage (Pc. The land accuracy in clear-sky (Lc) or in all-sky (Sa) is also defined in a similar way. the IO errors are relatively small and almost have no change in the snow cover period and snow-free period among those products. Those reductions of misclassification indicate the improvement in snow detection from single type (Terra or Aqua). which indicate that Terra products have less bias than Aqua products. Results 4. S means September and O means October. Table 3 presents the mean IU and IO errors against ground observations at both monthly and annual basis. daily combined and adjacent temporal deduction products are presented in Fig. in situ observations. Daily Terra and Aqua combined products (MODISDC) reduce monthly cloud coverage by 5–14%.1666 Y. d. Gao et al. and to adjacent temporal deduction products especially in transitional period. the IU and IO values of Terra products are less than those of the Aqua products (Table 3). and f represents the total data for all-sky. A) and snow coverage (Ps. Overall. so it can reflect the overall performance of a product. Those changes are mostly obvious in the cold season. Image: snow Ground: snow Ground: no snow a c Image: no snow b d Image: cloud e f a.0%. and to 27. The comparison of four different products indicate that IU errors have a decreasing trend from Aqua or Terra products. For example. 5B shows that monthly snow retrievals of these four products change in reverse trend. b. and adjacent temporal deduction (MODISATD) snow cover maps.1%. MODIS daily combination and adjacent temporal deduction The cloud and snow coverage (%) monthly distribution of Terra. The distribution of IU and IO errors shows a typical seasonal pattern of low values in snow-covered period (December–March) and snowfree period (July–September) and of high values in transitional periods (October−November. and d represents the total data under clear-sky. and April– June. respectively. They are expressed as below (in percentage): Sc = Sa = a a+b a a+b+e b a+b b : a+b+e ð7Þ ð8Þ ð9Þ ð10Þ Fig. 5A shows that the cloud coverage of standard daily MODIS products (MOD10A1 and MYD10A1) is high especially in winter months and that MOD10A1 (Terra) has slightly lower cloud coverage than MYD10A1 (Aqua). The major IO errors come from those transitional months. Fig. d. c. and so on. in May the mean IU error was reduced from 34. the IU errors of all four products are larger than the corresponding IO errors except in August and September.7% to 28. and the sum of a. The adjacent temporal deduction products (MODISATD) reduce monthly cloud coverage by another 6–9%. Lc = La = 4. snow accumulating period. and to adjacent temporal deduction products. 5. Aqua. to daily combined. The overall accuracy in all-sky condition considers both the classification accuracy and retrieval rate. The overall accuracy in clear-sky (Oc) or in all-sky (Oa) condition is respectively defined as the sum of matched number of station-pixels for snow and for no snow divided by the total number of stationpixels in clear-sky condition or in all-sky condition. e and f represent number of station-pixels in each particular classification category. one in clear-sky condition and the other in all-sky condition. and to the adjacent temporal deduction products (MODISATD). As compared with the IU errors.4%. b. / Remote Sensing of Environment 114 (2010) 1662–1675 Table 2 Confusion matrix for remote sensing image vs.1. c. The three year in situ observations at 244 SNOTEL stations are used to verify these snow cover products. All of these indicate that daily combination and adjacent temporal deduction are efficient to reduce cloud coverage and reveal the real snow coverage especially in cold season. c. both expressed as below (in percentage): Oc = a+d a+b+c+d a+d : a+b+c+d+e+f ð5Þ ð6Þ Oa = The sum of a. 30. e. while almost no change happens from June to September since there is almost no snow cover in those months. b. The snow coverage increases from the standard daily MODIS products (MOD10A1 and MYD10A1). daily combined (MODISDC). B) of the Terra MODIS (MOD10A1). Fig. Overall.

4. and continuously is increased up to 39.9 0.8 13.0 1.9%.3 15.9 17. and from 6-day to 8-day are 17. the mean annual increase of snow coverage (3.4 1667 IU Sep.0 11.2 28.4 15.1% for the 8-day combined map. it is hard to believe that the MODIS sensor in Aqua has similar and consistent performance as the MODIS sensor in Terra for all other months. daily combined and adjacent temporal deduction snow cover maps for the period of 2006–2008 hydrological years.1 12.2% for 6-day and 31.1 8.6 0.9% for 8-day products.1% for 4-day. while the snow coverage is increased to 24.0 2.9 2. Jul.2.4 0.7 4.4 6.0 2. The three columns on the left show the date.1 1.8 0. while. by fixed-day combination and flexible multi-day combination. the greatest cloud reduction and snow increase are from 2-day to 4-day. from 4-day to 6day. These excess snow increases might be due to the accumulation of snow coverage at different locations in different days.1 0. The seasonal variability of IU and IO errors is again typically low in stable periods (snow-covered or snow-free) and high in transitional periods.1 12.1 30. the magnitudes of Fig. This value is much larger than the 3.8 3. Aug. Aqua.1 0. B) of two flexible multi-day combined snow cover maps (MODISMC4 MODISMC8) as compared with that of the fixed 2-. This provides a strong justification for the flexible multi-day combined approach. This means that the magnitude of those improvements slows down although the losses of temporal resolution are the same (2 days).2 0.8% of the MOD10A1. instead.7 3. Feb.0 7.9 15.to 8-day MODIS snow cover combined maps from October 26 to November 2. Multi-temporal combination of Terra and Aqua MODIS snow cover products Fig.9%. A) and snow coverage (Ps. 2007).1 9. and so does the monthly changes. 2007.1 0. those changes are relatively large. In the stable periods.2 0. the overestimation increases.0 17.1%. Therefore. For example. the cloud reductions from 2-day to 4-day.2 0.2% and 4.2 12.Y.5 Adjacent temporal deduction MODISATD IU 2.1% for 2-day combined.9 0. cloud coverage is reduced from 11.5 1.5 6. That's why the NASA/ MODIS standard 8-day products (MOD10A2/MYD10A2). lower the time resolution and bring more classification errors.8 27. An example of 2.6 1.0 5.4 1.7%). that is called maximum snow cover product of the 8 days. Table 4 summarizes the IU and IO errors of the fixed-day and flexible multi-day combined snow cover products against three year ground observations at 244 stations.7 13.4 1. 4-. 6-.7 1. / Remote Sensing of Environment 114 (2010) 1662–1675 Table 3 Monthly and annual mean image under-(IU) and overestimation (IO) errors (%) of Terra.0 16. 3 of them have cloud coverage over 10%. .3 1. 6-. respectively.7 1.7% on November 1 while the cloud coverage of this day is 4. Apr.6 5. except the first 2-day combination. except for October for the 3 years in a row.0 7.6 4. So. Among the 8 daily products. All of these suggest that persistent combinations indeed overestimate snow coverage.5%) eventually exceeds the decrease of cloud coverage (2. Obviously.5%. followed by 4day to 6-day.2 0.2 11.6 0.2 0. 28. the persistent combination should be stopped at an appropriate time to reduce the occurrence of snow coverage overestimation.2 0.0 13.4 8.5 0.6 1. May Jun.8 One exception in the Table 3 is the large October IO error (11. and 6-day to 8-day combinations. which is similar to those of daily products (Table 3).7% and the corresponding snow increases are 13.8 IO 3. Mean percentage (monthly) of cloud coverage (Pc. is known to overestimate snow coverage (Hall & Riggs.3 5.3 12.2 5. further explanation for this observation remains open. Mar. From 2-day to 8-day fixed combinations.0 2.6 0. Month Aqua MYD10A1 Terra MOD10A1 Daily combined MODISDC IU 2.9% of the first day to less than 1% for 2day combined (0 of most other combined products).1 0.2 0. 10. and 8-day combined results during 2006–2008 hydrological years. the other combinations are not needed since in those combinations almost no cloud coverage can be reduced while excess snow coverage is accumulated.2 0.5 IU 2.4 0. In the same period.4 12.2 0. cloud and snow percentage of the daily products.3 7. those changes are small. However.3 0. and 8-day and flexible multi-day combined snow cover products. some thresholds should be used for the useful combinations to remain and reduce the unneeded combinations.3 1. in January.1 2. It is clear that fixed-day combination progressively decreases cloud coverage and increases snow coverage as the increase of composite days. The largest IO error in October increases from 15.5 6.5 5.1 2.3 4.2 12.3 1.. 6.6 0. Oct.3 3.6 5.0 0.2 IO 3. Dec.8 2. the annual error and most of the monthly IU errors keep reducing while the IO errors keep increasing.7 IO 0. Annual 2.9%) of the MYD10A1.6 0.2% and 5. Thus. the three columns on the middle show those of persistent fixed-day combined maps. some unneeded combinations really exist.6 IO 3. This reveals that in persistent combinations. The maximum snow coverage is 20. particularly in the snow onset and melting periods. For the fixed-day combined maps.2 11. 6 shows the mean monthly distributions of cloud and snow coverage of fixed 2-. and the three columns on the right show those of flexible multi-day combined maps. Pu et al. One possible explanation is that the MODIS sensor in Aqua might not have been in a good condition in October.6 34. Jan.2 4. 4-. in the transitional periods. 11. This indicates that along with the increase of temporal windows. Gao et al.9 0.4 1. Comparing the magnitude of cloud coverage decrease and snow coverage increase between 6-day and 8-day combined products. These unneeded combinations do not reduce the cloud coverage.1 3.1 0.7 12.9 5.7% for 2-day to 22.6 0. Nov. is presented in Table 5.

4.1668 Y.6 19.3. 29–30 Oct. daily MODISATD_AE.0 7.6 2. The direct comparison between the standard daily AMSR-E products and MODIS products (Table 3) shows that both have higher IU than IO errors. 39 MODIS4DC and 61 MODISMC8 vs. 2008 Oct. and 7% from MODISMC8.1 10.5 18.7 15. we can map and track two snow melting and one snow accumulating events.0 Ps (%) 16.0 4.1 7.1 25.9% in July to 75.6 3. After multi-sensor combinations.6%.3 3.2 0.0 2.2 0.5%) and similar.8 8. Annual 2.9 4.9 2. Fig.0 2.1 10.5 1.0 0.9%) in December and the monthly snow percentage ranges from 0.2 17.6 6.8 11.7 9. 0. The comparison between multi-sensor combined products and its corresponding MODIS products (MODISATD_AE vs.1% (3. Nov.3 3. Passive microwave remote sensing has the ability to penetrate clouds. 2008 Oct.2% (Table 5).0 2. 2008 Nov.9 1. while the cloud coverage of these flexible multi-day combined products in snow-covered period is lower than that of MODIS6DC.7 3.8 5.3 0.6 20.9 36.2 1.3 19. The first snow melting event happened between the late afternoon of October 27 and October 28 (raw cloud coverage of those 2 days were 1.4 6. The mean snow coverage of daily AMSR-E ranges from 0.4 4.9 3.0 0.0 2. 2008 Nov.7 3.9 3. The curves of cloud and snow coverage for MODISMC8 are similar to those of the fixed 6-day combined products (MODIS6DC).4 20.2 0.7 15.2 0. 2008. The cloud coverage of MODISMC4 (MODISMC8) ranges from 3.6 1.7 9.1 0. MODISATD Oct.9 1.6 3.7 28.6 For example.2 1.1 1. and 3.6 9. and multi-day MODISMC4_AE and MODISMC8_AE.3 2.2 0.5 10.2 9. please refer to Table 7). MODISMC8_AE product has the smallest cloud replacement.7% from MODISMC8.7 0.4 1.4% from MODISATD.1 IO 8.4 5.6 10. The curves of cloud and snow coverage for MODISMC4 are similar to those of the fixed 4-day combined products (MODIS4DC).7 4. respectively.0 6.6 5.0 3. Apr. Therefore. increase the annual IU errors by 11.9 31. 1.2 0.2 .1 19.2 IO 17.8 2.7 39. Jan.5% and 15.6 IO 7.0 6.1 IO 13.4 19. 31. Month Fixed-day combined 2-day IU Sep.4% in January. 26–27 Oct.6 8.3 0.3%.9 4.2 0. 2008 Pc (%) 11. respectively.1 3. in which snow cover increased to 20. The IO errors of both MODIS and AMSR-E are small (less than 2. 28.4 6. Then a big snowfall event on October 29 and 30 (raw cloud coverage were 19. 2008 Oct. 7. MODISATD. MODISMC4_AE vs.4 1.1 0. and provide a better monitoring of the temporal variability of snow coverage.4% and 6.5 4-day IU 2.1 18.1 38. daily MODISATD_AE. The IU and IO errors of flexible multi-day combined products (MODISMC4 and MODISMC8) have a similar pattern as the fixed 4-day combined products. the three new products.4 3. thus the largest increase of the IU errors.9 15.8 22. 61 MODIS6DC). Feb. This should enable a more reliable identification of the onset and ending date of snow cover. From them. and 0. 31.7 5.4% (0.9 1.4 15.5 1.5 3. 2008 Pc (%) 0. May.0 3.3 6-day IU 2.3%. MODIS flexible multi-day combined snow cover products have more images available during the whole hydrological year (160 MODISMC4 vs.1%) in July to 27.1 MODISMC8 IU 2.6 11.8 0.6 0.4 2. Multi-sensor combination of MODIS and AMSR-E The distributions of mean monthly snow coverage of AMSR-E and multi-sensor combined snow cover products are presented in Fig. 2.6 20.1 7.7 Multi-day combined 8-day IU 2. 6. also even in the transitional months (68 MODISMC4 vs.6 0. The degree of monthly and annually underestimation error (IU) of the AMSR-E products is much higher than that of the MODIS products since the spatial resolution of AMSR-E is coarse.3 22.8 5.1 0. 6 also shows the mean monthly cloud and snow coverage of flexible multi-day combined snow cover products MODISMC4 and MODISMC8.2 3. 2008 Oct.6 4. Table 6 shows the classification errors of AMSR-E and multi-sensor combined snow cover products. MODISMC4 and MODISMC8_AE vs. 91 MODIS4DC and 141 MODISMC8 vs.1% and cloud coverage ranging from 0.6% to 24.9 Ps (%) 16.8 2. Gao et al.4 11. so the AMSR-E product has no cloud obscuration. 27 MODIS6DC.9 10.7 26.0 19. 30.0 MODISMC4 IU 2. 14.4 6. respectively).1 4.4 2.5 0. and multi-day MODISMC4_AE and MODISMC8_AE.6%.2 1.7 8.3 0.4% to 9.8 0. 2.8 6.6 9.0 0.0 IO 5.6 31.9 24.0 4. and 2.3%.2 8.7 1.3 11. 5.5 1.3%.6 4.1% from MODISMC4. 2008 Oct.3 0.3%) in January.4% and remained until snow melted again on November 2.1 7. Jun. 27. 29. remove the annual cloud coverage by 36. our flexible multi-day combinations generated 6 images over the 8-day period with snow coverage ranging from 11.7 6.0 0.4 2.4 7.5 6.7 8.1 7.8 10.4 0.1% to 11.4 8.2% from MODISATD. Mar.2 0. 28 Oct.6 10.6 0.7%.2 0.0 0.7 IO 9.2 Combined 1 day 2 days 3 days 4 days 5 days 6 days 7 days 8 days Pc (%) 11.4 4.3 4.3% from MODISMC4.7% (11.4 2.4 1.1 2.5 1.2 8. Oct. / Remote Sensing of Environment 114 (2010) 1662–1675 Table 4 Monthly and annual mean image under-(IU) and overestimation (IO) errors (%) of the multi-temporal (fixed-day and flexible multi-day) combined snow cover products for the period of 2006–2008 hydrological years. their fixed-day combined and flexible multi-day combined snow cover maps. The three new products. Dec.7 2.1 1.2 4.6 0.6%. that reduced the snow coverage from 24.2 0. and their seasonal variability is also low in stable periods and high in transitional periods (Table 4).6 1.0 6. 1. thus the smallest increase of the IU errors.7 7.6 20. In contrast.2 8.7 0.5 0.3 23.5 15.7 1.9 Ps (%) 24.4 1. and increase the annual snow cover retrieval by 13. The MODISATD_AE product has the largest cloud replacement. 2008 Nov. Aug. 26.4%.4 8.4%) in July to 56.1 5.2% (70.3 9.0 5. multi-sensor combinations do cause more uncertainty due to the coarse spatial resolution of Table 5 Percentage of cloud coverage (Pc) and snow coverage (Ps) of daily MODIS adjacent temporal deduction maps.3 0.0 5.1 24.2 0.1 MODISMC Oct. and decrease the annual IO errors by 0. respectively) followed.9 0.1 11.0 4. 2008 Nov. Jul.2 0. MODISMC8) indicates that the multi-sensor combination tends to increase the IU errors and decrease IO errors to different degrees.0 8.

so does the monthly accuracy.4 8.9 5. Jan. Similar to the change trend of the annual overall accuracy and cloud reduction. the more uncertainty (i. For the multi-day (fixed and flexible) combination products.2% and 89. however. This is at the cost of the accuracy.4 14.2 20. during the other 7 months.2 0.0 3. as the temporal combinations applied. respectively.1%.0 2. the Terra snow cover products have slightly higher accuracy and lower cloud coverage than the Aqua products in every month (squares with different gray scale).1 0. 6-.3 5.4 46.0 27. 8).5 26.0 12.2%.4% with respect to MODIS Terra product.1 0.3 22. 4.9 0.0 1.4%.3 32. The snow accuracies are very high in snow-covered months (November to April). Jun. Meanwhile its cloud coverage decreases by 7. Oct.2%. but also achieve higher accuracy in every month.1 24.5 1.3%.Jul. .3% and 7. with the monthly accuracy progressively increasing. the adjacent temporal deduction products provide a 0. September and October. 9). However. / Remote Sensing of Environment 114 (2010) 1662–1675 1669 Fig. Dec. Table 6 Monthly and annual mean image under-(IU) and overestimation (IO) errors (%) of AMSR-E and multi-sensor combined snow cover maps for the period of 2006–2008 hydrological years.6 Multi-sensor combined MODISATD_AE IU 4.0 6.7 1. Table 7 also shows the snow and land accuracies in clear-sky and all-sky conditions among different snow cover maps.3 19.7 10.Y. June.2% and 1.2 1. as the cloud coverage reduces (Fig.2 .1 0. In the daily products.4 21. due to lower land accuracies (not shown).1 0. however. Accuracies in clear-sky condition of different snow cover products and its trade off for cloud reduction The accuracies of different snow cover products are quantitatively evaluated against three year observations at 244 stations.5 2.4 49. the cloud coverage decreases and the accuracy also decreases continuously. In contrast.3 4. 39. from both multi-sensor products and their MODIS counterparts are similar (see Tables 4 and 6). This means that fixed-day combinations do bring uncertainty and more combination will bring larger uncertainty. Apr. as compared with their MODIS counterparts. The IO errors.e. IU error) is introduced to the multi-sensor combined products. which decreases by 0. the cloud coverage reduces.7 1.1 0.6 5. However the cloud coverage decreases to 14.6 1. The higher snow accuracies (N 80%) in November do not result in higher overall accuracies in this month (Fig. the cloud coverage reduces continuously.2 18. respectively.5 19. Table 7 summarizes the annual overall accuracies in clear-sky condition and mean cloud percentage of different snow cover products.8 0. 2.8 7. Feb. Differing from the annual change trend.8 1. during 2006–2008 hydrological years.3 7.7 1. 1. 8). while much larger in all-sky condition. Aug. which can meet most of user requirements. increase trend. do not affect the overall accuracy for those 2 months (98% or higher) (Fig.9 31.7 13.7%.0 0. In other words. The cloud coverage decreases by 7. More details about the tradeoff between the overall accuracies in clear-sky condition (Oc) and the cloud reduction of different snow cover maps are illustrated in Fig. Lower snow accuracies (30–80%) in May.2 15.7 15. In the clearsky condition.4%.8 16. 4-.1 15. Terra and Aqua combined products give a similar accuracy with a loss of 0.5 0. the land accuracies are larger than those of snow accuracies in clear-sky condition.6 2.2 0. The adjacent temporal deduction products (star) not only reduce the cloud coverage with respect to daily combined products.2%.9 13.1 0.1 1.8 IO 8. the monthly snow accuracy (Sc) in clear-sky condition in all months shows only one trend. In June. Gao et al. 7. The 2-. The flexible multiday combined snow cover products MODISMC4 and MODISMC8 have an overall accuracy of 89.1 0.9 28. Mar.8 2.1 13.3% higher accuracy with respect to daily combined products.8 0.6%.7% compared with MODIS Terra product.2%.4. Nov.1 1. July. and 2.1 3.6 30. August.0 0. which is slightly lower than Terra snow products by 1. the more cloud coverage is replaced.1 9. and 48.8 1.0 11.0 43. respectively. the differences between those accuracies in clear-sky or in all-sky condition are smaller. 45.4 4.1 12. Annual 5. in the period of 2006–2008 hydrological years. AMSR-E products.5 40.6 17.5 22.7 IO 6. monthly accuracies for daily.0 4.1 19. Month AMSR-E AE_Dysno IU Sep.1 0. The very low snow accuracies (∼ 20% or less) in July and August. 6.8 MODISMC4_AE IU 3.9 0.7%.1 0. and 8-day combined products reduce the cloud coverage by 25. and are even very close to 100% for all products from December to March.3 34.6 1. The patterns of flexible multi-day combined products MODISMC4 and MODISMC8 (triangles with different gray scale) are respectively similar with that of 4-day combined and 6-day combined products except in May.9 3.4% while the cloud coverage is high at 52.0 20. 8.9 7.3 0. since numbers of those misclassified pixels are very small.2 3. For the fixed-day combined products. In contrast to varied seasonal patterns of monthly overall accuracy (Oc) presented in Fig.3 1.5 MODISMC8_AE IU 3. 4-.0 0. with the increase of the temporal window. the original Terra snow cover product MOD10A1 has the highest accuracy of 90. 8.3 IO 2. and October indeed affect the overall accuracies in those 4 months (Fig.6 20.0 0.0 IO 2. 2-. Mean percentage (monthly) of snow coverage (Ps) of AMSR-E (AE_DySno) and multi-sensor combined daily (MODISATD_AE) and multi-day (MODISMC4_AE and MODISMC8_AE) snow cover maps. September.1 18.6 1.2 0. May. 8).8 2.7%.9%.1 12.0 2.3 3.and 8-day combined products (cross and circles with different gray scale) show some seasonal patterns.

and MODISATD).8 73. Trade off between the monthly overall accuracy in clear-sky condition Oc (Eq. The flexible multiday combined products MODISMC4 and MODISMC8. respectively at the mean loss of 2. This is due to the advantages of the effectiveness of MODIS high spatial resolution and AMSR-E cloud penetration ability.9 Multi-day combined AMSR-E Multi-sensor combined 4.5.8 86.5 95.4 84.2 / / / / Sa (%) 25.0 80. Fig.9%.and 8-day combined products with spatial resolution of 500 m.0 0.6 days temporal resolution. land accuracies in clear-sky (Lc) and all-sky conditions (La).6 51.8 81.0%.3 76. (1)) of different snow cover products in 2006–2008 hydrological years (S is September.1 90.2 88.0 12.0 0.9 92. Fig.0 34. 10 shows the variability of monthly overall accuracies in all-sky condition vs.7 79.8 89.8% and 86. 11 shows the regional variability of the overall accuracy in allsky condition for different daily snow cover products.2 56. mean cloud percentages (Pc).1 27.4%. give the accuracy of 72.4 Lc (%) 98.1 79.1 89.9 64.5 89.3 73.0 89.0% and 79.0 31. has an highest overall accuracy of 79.3 89. in which the overall accuracy increases as the cloud coverage decreases.5 3.2 87.0 88.4 83.5 60. the number of images in one hydrological year (transition period: October–November and April–June).6 91.6%.8 66. in the period of 2006–2008 hydrological years.6 95. O is October.0 79.9%.5 37. / Remote Sensing of Environment 114 (2010) 1662–1675 Table 7 Comparisons of overall accuracies in clear-sky (Oc) and all-sky conditions (Oa).to 8-day. This is at the expense of decreasing temporal resolution from 2.9 82.0 0.6 72. (5)) and cloud coverage Pc (Eq.7 67.2 83.6 91. to 79. The accuracies are calculated separately based on the four elevation zones of SNOTEL Fig. among different snow cover maps.0 Total images 363 365 365 365 183 91 61 46 160 141 364 365 160 141 (151) (152) (152) (152) ( 76) ( 39) ( 27) ( 21) ( 68) ( 61) (152) (152) ( 68) ( 61) Sc (%) 79.1 36. MODISDC. Oc (%) Terra Aqua Daily combined Adjacent temporal deduction Fixed-day combination MOD10A1 MYD10A1 MODISDC MODISATD 2 days 4 days 6 days 8 days MODISMC4 MODISMC8 AE_DySno MODISATD_AE MODISMC4_AE MODISMC8_AE 90.7 90.6 87.8 93.4 / / / / La (%) 60. the accuracy increases from 60.8 14.9 84.9 83.). with a spatial resolution of 500 m.5 89.5 6.4 74.9 79. Table 7 also shows the annual overall accuracies in all-sky conditions and the mean number of images within one hydrological year and in those transitional months (October–November and April–June).4 77. Gao et al. 4-.2%. and much larger than accuracies of the other four MODIS products (MOD10A1.6 78. to 73. cloud percentage for different snow maps.0 83.2 83. 6. 8.4 20. The characteristics of the monthly change are very similar to the annual situations shown in Table 7.7 66. and etc.8%. For all daily products.5 96.8 95.0 Oa (%) 40.8 83.5 76. the multi-sensor combined snow product (MODISATD_AE).2 44. MYD10A1.0 47. and to 82.3 45.1670 Y.7 82.3 / / / / Pc (%) 52. For the fixed 2-.1% of the AMSR-E snow product.2 0.3 7. Regional variability of accuracy in all-sky condition and trade off between temporal resolution and spatial resolution The overall accuracies in all-sky condition can evaluate the real performance of different snow products in applications such as hydrological modeling. snow accuracies in clear-sky (Sc) and all-sky conditions (Sa).5 89. which is slightly larger than the 74.7 54. .2 95.5 58.2 89.6 89. while they lost certain spatial resolution.0 85. with the pixel size of 500 m.5 83. The multi-day multi-sensor combined products MODISMC4_AE and MODISMC8_AE also have higher accuracies of 83.3 to 2.4 88.

48 stations in elevation zone 0–1500 m. Overall. Namely.). Remote sensing images really have difficulty to detect scattered (minority) objects in transitional periods. (6)) and cloud coverage Pc (Eq. O is October. Trade off between the monthly snow accuracy in clear-sky condition Sc (Eq. The monthly overall accuracy in all-sky condition Oa (Eq. or minority Fig. / Remote Sensing of Environment 114 (2010) 1662–1675 1671 Fig. and etc. 10. and 55 in elevation zone over 2500 m. O is October. stations. the MODIS products (two left panels) have much lower accuracies from November to May as compared to those from June to October. (1)) of different snow cover products in 2006–2008 hydrological years (S is September. 9. . 77 in elevation 2000–2500 m. and etc.Y.). AMSR-E products (top right) and daily AMSR-E and MODIS combined products (bottom right) have higher accuracies (over 60%). (7)) and cloud coverage Pc (Eq. This is mainly due to the scale difference from point (SNOTEL station) to pixel (MODIS pixel). Gao et al. except in transitional periods. (1)) of different snow cover products in 2006–2008 hydrological years (S is September. 64 in elevation zone 1500–2000 m.

12. while the lowest accuracy occurs at the lowest elevation zone (less than 1500 m).1672 Y. For MODIS products.). the accuracies show no consistent relationship to elevation. but not necessary in the corresponding image pixel scale. The highest accuracy occurs at the highest mountain zone (above 2500 m). 11. 2010). For all products in the period of November to April. (6)) of different daily snow cover products in different elevation zones (S is September. Gao et al. (6)) of different multi-day snow cover products in different elevation zones (S is September. and etc. and accuracy increases as the combination day increases. the Terra MODIS product has slightly high accuracies than Aqua MODIS product in all months at all elevations. O is October. the accuracies of every product are relatively high. that snow cover at lower elevation is more changeable and can be a scattered minority object that can be seen in a station. In other months. the only difference in Fig. In July. The most obvious increase for all the MODIS fixed and flexible multi-day combination products is achieved during the period from November to May when the accuracies are generally Fig. Overall. . The monthly overall accuracy in all-sky condition Oa (Eq. August and September (snow-free period). the daily combined Terra and Aqua MODIS (MODISDC) product has higher accuracy than either Terra MODIS (MOD10A1) or Aqua MODIS (MYD10A1) in all months at all elevations. surface objects are not big enough to be resolved in coarser resolution images (Gao et al. Similar to Fig. 12 is that it shows the accuracies for all eight multi-day products. 11. O is October. / Remote Sensing of Environment 114 (2010) 1662–1675 Fig. and etc..). the accuracies clearly depend on elevation. multi-day products have higher accuracies than the daily products. although the magnitude of increase decreases. and the adjacent temporal deduction product (MODISATD) has a higher accuracy than the MODISDC in all months at all elevations. The monthly overall accuracy in all-sky condition Oa (Eq. This is the same reason as discussed for the transitional months.

e. This is also similar to what is seen for the daily products (Fig. In all months except December. Meanwhile. This means that. In addition.5%. 46% for the spatial filter. are very similar and very high from November to April. with respect to the daily combined snow cover map.9%. One is the different mapping accuracies and physiographic characteristics of the SNOTEL stations.8% (8-day combined). 94. Gao et al.8%. The adjacent temporal deduction method uses the information of two temporal adjacent images to deduce cloud pixels on the current day image. Therefore the persistent combination should be stopped at an appropriate time for the useful combinations to remain and reduce the unneeded combinations. the cloud coverage decreases and the snow coverage increases progressively. maximum cloud percentage (P) and maximum composite days (N). In this study. the overall accuracy for snow-covered months are actually very high over 95%. so does the accuracy as progressively more data are merged. 6.6 days/image. this method reduces the IU/IO errors and increases the overall accuracies in all-sky condition to 51. reduces from June to October (not much change in July and August) as cloud decrease. In particular.. while the overall accuracy in clear-sky condition slightly decreases from 90. However. 9).7% over MYD10A1.5%.to 8-day consecutive images to produce maximum snow coverage and minimum cloud coverage. misclassification error and overall accuracy against three year observations at 244 SNOTEL stations. 12. This would have contributed in reducing the overall mapping accuracy of the study area. particularly in the snow onset and melting periods (transition periods). 2009. there is no way that the method can be used in operational context. This suggests that. 2008a. and the corresponding overall accuracies are 95. respectively. Interestingly. 2007). it is by no means that one has to use it before any multi-day or multi-sensor combinations can be performed. the accuracy in all-sky condition for the fixed 2-.4% reduction in cloud coverage.0% to 89. and to 3. 34% for the 1day temporal filter. 4-. from this group of elevation/SNOTEL stations. Wang et al. ∼ 20% of them are located at elevations lower than 1500 m. 8. This is due to the larger mapping errors in those months as discussed earlier (Tables 3. If one pixel in the current image is cloud.2% and 87. Discussion and conclusions The overall accuracies in clear-sky condition (88–90%) of daily Terra and/or Aqua MODIS products in the area of study (Table 7) are slightly lower than the ∼ 93% from other areas (Hall & Riggs. to 79. In particular. since many other studies used 1–4 cm of snow depth to define snow or no snow from in situ measurements (Liang et al. 2006). Tekeli et al. remote sensing. i. in the 3 years. Therefore. 2009. The overall performances of different snow cover maps are quantitatively evaluated through cloud and snow percentages.5%. although from 2-day to 4-day. Compared with fixed-day combinations. 11). However.. Last but not the least. for the threshold N = 4. Those low values from the five transitional months have lowered the overall accuracy for the area. etc. 94. As the temporal window gets larger. the variability of monthly overall accuracy (in clear-sky condition) shows a typically seasonal pattern. Wang & Xie.2%. 89. Among the 244 stations. flexible multi-day combination and multi-sensor combination are examined and applied in the Pacific Northwest USA. the procedures were terminated by P for 404 times and by N for 19 times. the magnitude of cloud reduction and accuracy increase slows down. however.. China. The results reveal that combination reduces the mean annual cloud coverage over the study area from 37. respectively. 2009. and June).. The primary objective of this study is to examine different approaches in reducing cloud obscuration and to generate products that satisfy the requirements for different applications. 2009. Terra and Aqua MODIS combination reduces the cloud obscuration by 11. 2008. April. 6. of course.9%. (2009) in cases of Colorado Plateau and in Northern Xinjiang. all products exhibit better accuracy at higher elevation. from 4-day to 6-day and from 6-day to 8-day.1%. A similar temporal filter was tested by Parajka and Blöschl (2008) over Austria and verifies against daily snow depth observations at 754 climate stations.1% (MODIS daily adjacent temporal deduction snow cover maps) to 27. the overall accuracies actually increase due to the progressive combinations to reduce cloud coverage. two thresholds. 2005. Our monthly and annual results indicate that some unneeded combinations really exist in the persistent combinations. The snow accuracies in clear-sky for all products (Fig. the threshold value of 0. it should be noticed that although the adjacent temporal deduction shows the advantage and improvement on snow mapping accuracy. adjacent temporal deduction. In this study. November. As shown in Figs. Maurer et al.3 and 2.. Yang et al. the procedures were terminated by P for 366 times and by N for 114 times. for the threshold N = 8.0% and 79. from 2. although they are much lower in other months or even less than 20% in July and August. respectively. which is consistent with Xie et al.4% and 92. Fixed-day combined approaches merge 2. 94. Thus the accuracy in all-sky condition increases from 40%/34. / Remote Sensing of Environment 114 (2010) 1662–1675 1673 lower than those of other months in the MODIS daily products (Fig. i. but increases from November to May (Fig. most of the procedures were terminated by the threshold P (10%). the losses of temporal resolution are all the same 2 days. Wang et al. there were 19 cases of continuous 8 days or 114 times of continuous 4 days. the values are much lower.and 8-day combined products increases from 60. 2003. Also during this period. In the 1–4 cm snow depth range. 8). and 4% for the 7-day temporal filter.8%. Its spatial resolution is 500 m and its mean temporal resolution is 2. the adjacent temporal deduction is an efficient method to reduce cloud obscuration and improve the overall accuracy.5% (4-day combined).5 mm) of snow depth in this study used to define snow/no snow from SNOTEL station would also contribute to the lower overall accuracy of MODIS products. the cloud coverage decreases.4% to 45. would achieve high overall accuracy. May. Their results demonstrate that 63% cloud coverage of the Aqua images is reduced to 52% for combined Aqua–Terra images. since remote sensing has difficulty to map thin and marginal snow covers as discussed before. and be more suitable to map and .9%.6%. snow N no snow (land) N clouds. fixed-day combination.). 2008. while the excess snow indeed happens since the accumulation of snow coverage at different locations in different days.5% (6-day combined). however. and 6). This could be attributed to several reasons. Xie et al. For example.1 in (or 2.e.7% over MOD10A1 and 7. in the snowfall and most snow-covered months. 2010). this product should enable a more reliable identification of the onset and ending date of snow cover. 11 and 12. 88. to 73. Various approaches including daily combination. there is a big difference in overall accuracy from month to month or from season to season.Y. 11).. MODIS shows the lowest overall accuracy since the scale difference from SNOTEL point to MODIS pixel (Gao et al. and to 82. an image in a previous day can always been used for this approach. Second.4%.6%. Xie et al. In our case of three year data for the study area. 5. while land (snow) is shown in the corresponding pixel in both preceding and following images. as shown in Fig. with mean cloud cover of the area over 10%. These two thresholds are user-defined parameters that can be changed depending on the application.to 8-day combinations. while for those transitional months (October. It is very clear that the accuracies in the two transitional periods (October–November and April–June) are lower than other months in all multi-day combination products.0%. Similar results have been reported in many other researches (Parajka & Blöschl. are used to control the flexible temporal windows and generate flexible multi-day combined snow cover products. 4..0% (2-day combined). this cloud pixel of the current image is then assigned as land (snow).. Daily combination merges two MODIS snow products on a pixel basis according to the priority principle.. respectively. The accuracy in all-sky condition of flexible 4-day and 8-day combined products is 72. Application of this method results in an additional 7.

Alaska area. 2) Among the fixed-day compositions. However. Critical reviews and constructional comments from five anonymous reviewers to improve the quality of this manuscript are greatly appreciated.3%. while significantly increasing the overall performance since its 100% retrieval rate. (2) The annual IU errors of fixed 2. Overall. (2004) validated MODIS MOD10A1 (500 m) and NOAA GOES+ SSM/I (4 km) products for the period between January and June 2001 against 2000 surface snow depth measurements across Canada. / Remote Sensing of Environment 114 (2010) 1662–1675 track the temporal variability of snow coverage. which are similar to the daily products. and the IU errors are always slightly larger than IO errors for all fixedday combined products. Simic et al. Acknowledgements This work was in part supported by the U. Their results indicate that the accuracy of multi-sensor combined products has the snow accuracy in all-sky condition of 86% with respect to 31% (MOD10A1).S. but not as comprehensive as what we tested here. They found that snow identification from the combined products outperformed the ones based on a single satellite. K. but with more images available (one image per 2. These findings have important implications as indicated below. Its pixel size is 500 m and its mean temporal resolution is 2. multi-sensor combined products can eliminate the effects of cloud obscuration. MODIS and AMSR-E) combined approaches are robust techniques which cannot only reduce the cloud coverage. and 3) While flexible multi-day combination is found to outperform the fixed-day combined products in reducing cloud reduction and improving image accuracy. The difference between them is that multi-day combined methods sacrifice some temporal resolution while multi-sensor methods sacrifice some spatial resolution. Advanced Water Resource. 2-. which match better with the fixed 6-day (4-day) combined products. D. In the snow-covered months. but provides more available images. This is because these blended snow cover products utilized advantages of the effectiveness of MODIS high spatial resolution and AMSR-E cloud penetration ability.6 days). Similar evaluations for multi-sensor combined products have been performed in diverse regions and for various sensors. Gao et al. 45% (MYD10A1) and 85% (AMSR-E). The overall accuracy in all-sky condition of multi-day and multi-sensor combined products MODISMC4_AE and MODISMC8_AE is 83. but retains some cloud coverage. The MODIS and NOAA products show similar monthly agreements ranging from 80% to 100%.or subbasin-scale applications. References Andreadis. although the monthly percentages of cloud and snow cover change in the MODISMC8 (MODISMC4). 2.8% and 86. (2008b) produces new daily snow cover products through combining information of Terra MODIS daily snow cover product (MOD10A1) (the Aqua MODIS snow cover products not included) and AMSR-E daily SWE product (AE_DySno) in Northern Xinjiang. the flexible 8-day combination can reduce more cloud (less than 10%). P. The assessment shows that the overall accuracy of daily AMSR-E snow cover products. (2006). an overall overestimation of snow cover only occurs with longer composite days (6–8 days). our results demonstrate that flexible multi-day and multisensor (i. while significantly overestimates the snow cover in transitional months. The choice of different approaches will depend on the requirement of specific applications and the meteorological conditions of various regions. (3) The IU and IO errors of flexible multi-day combined products (MODISMC4 and MODISMC8) have a similar pattern to the fixed 4-day combined products. The multi-day combined map with higher spatial resolution has higher accuracy in clear-sky condition. the flexible 4day combination may leave more cloud cover in the winter months (over 10%).and 4-day combined products are larger than the IO errors. The multi-sensor combined approach utilizes the information from AMSR-E to update the remaining cloud pixels in the MODIS daily or multi-day combined snow cover maps. M. but also remain in good agreement with ground observations. the fixed 4-day combination products seem to be a good compromise to maintain the balance of temporal resolution and a realistic estimation of the snow cover extent. (2010) generated new daily snow cover products utilizing Terra/Aqua MODIS and AMSR-E snow products in Fairbanks and Upper Susitna Valley.e. and 4-day). Therefore. (2000) develops a method of automated monitoring snow cover over North America using GOES and SSM/I and validates their products against in situ land surface observations during the winter season of 1998/1999. Assimilating remotely sensed snow observations into a macroscale hydrology model. but lose some spatial resolution. these errors are actually small. and (4) the MODISMC4 and MODISMC8 have the same overall accuracy in clear-sky (89. The low retrieval rate depletes the effectiveness of the MODIS's high spatial resolution and decreases the overall accuracy of MODIS Terra in all-sky condition to only ∼ 40% (Table 7). Provisions of SNOTEL data by US Department of Agriculture's National Resources Conservation Service (NRCS) and MODIS and AMSR-E products by NASA EOS and NSIDC are sincerely acknowledged.. while the retrieval rates are 40% for MODIS and 100% for NOAA. NASA project (#NNX08AQ87G). The most interesting outcomes from this study are summarized as: (1) The primary errors of image underestimation (IU) and overestimation (IO) occur during transition months.4%) in clearsky condition of daily Terra MODIS snow cover products. it actually slightly underestimates snow cover in those snow-covered months. but may cause more snow overestimation in transitional months. The multi-sensor approach is the practical method that yields cloud-free snow cover products. 1) Although the standard MODIS 8-day product represents the maximum snow cover of the 8-day period. 872−886. Meanwhile.1%). . Daily MODIS and AMSR-E combined products with the pixel size of 500 m have the highest overall accuracy in all-sky condition of 79. This (74. respectively. Liang et al. which reflects the effectiveness of the MODIS's high spatial resolution.1%) is lower than the overall accuracy (90.1674 Y. while overall accuracy of MODISMC4 in all weather conditions is slightly less than that of MODISMC8 (72.3 and 2.. respectively. All of these studies indicate that multi-sensor combination is a useful method to improve the snow retrieval rate. Romanov et al. it is argued that such a flexible combination method is relevant to maintain the balance of temporal resolution and realistic estimation of the snow cover extent. 80. For example. 29.3% vs. The use of 4 and/or 8 days combinations as a threshold remains to be determined by the user as required for their specific application and/or study area. especially during snow accumulation period. with the coarse spatial resolution of 25 km.4%. but only for one hydrological year (October 2006 to September 2007). is 74.1% with the retrieval rate of 100%.3 days vs.4% (MOD10A1) to 75.6 days/image. China during the period from November 1 to March 31 of years 2002 to 2005. & Lettenmaier. The first author wants to thank the China Scholarship Council for funding her to study for 2 years (2007– 2009) at the University of Texas at San Antonio.2% among the daily snow cover products. Gao et al. Therefore the coarser resolution NOAA product may be sufficient for basin. the retrieval rate of Terra MODIS is below 50% due to high cloud coverage. The combination significantly increases the snow accuracy in all-sky condition from 33.2%). This indicates an overall underestimation of snow cover from those products (1-.

(2009). Journal of Applied Meteorology. Wang. X. Snow cover monitoring in the Northern Patagonia Icefield using MODIS satellite images (2000–2006).. Wang. (2009). Kirnbauer.. St.. Kim. D05113. Udnaes. 2002. & Xie. G. (2007).. L.. Carroll.. & McMahon.. Updated daily AMSR-E/Aqua daily L3 global snow water equivalent EASE-Grids V002. Y. Topographic control of snow distribution in an alpine watershed of eestern Canada inferred from spatially-filtered MODIS snow products.: Department of Mathematics and Computer Science. Updating a land surface model with MODIS derived snow cover. T. A. Wu. J. 1361−1373.. P. Robinson. & Houser. et al. 184−194. 214−231. B. Combining vegetation index and model inversion methods for the extraction of key vegetation biophysical parameters using Terra and Aqua MODIS reflectance data. (2008).. New Orleans. Feng. H. A.. & Easterling. T. Sorman. Spongberg. (2007). Foster. Romanov. (2009). R. (2005).. T. et al. 38(1).. P.. Huang.. Hydrological Processes. J.. and GOES + SSM/I snow cover products over Canada based on surface snow depth observations. Hydrological Processes. J. W. (2007). R. X. 679−689. & Jackson. Klein. doi:10. Trend identification in twentieth-century U. Shibata. Montesano. E. A new blended global snow product using visible. 162−176. Hydrology and Earth System Sciences. K. J.1029/2008JD011272. K. Zhou. Alfnes. Hall. Improving runoff modeling using satellite-derived snow cover area. St. & Qin. 21−32. & Riggs.. (2008). doi: 10. Grayson... Proceedings of the 64th Eastern Snow Conference. Tedesco. New methods for studying the spatiotemporal variation of snow cover based on combination products of MOIDS Terra and Aqua. (2008). (2009). Xie. 28 May–1 June 2007.. K. September to August 2005–2008. An assessment of the differences between three satellite snow cover mapping techniques. W. Distributed snowmelt simulations in an Alpine catchment. 105. Blended visible. R.. X. J. D. A. E. 033559. Global and Planetary Change.. Kawanishi.. L. Benko.1175/JTECH2017. Western. & Lettenmaier. M. 59−71. Users guide. Remote Sensing of Environment. 114. Water Resource Research. 1313−1334... John's. E. T. Brown. L. Hydrological Processes. J. Toward advanced daily cloud-free snow cover and snow water equivalent products from Terra–Aqua MODIS and Aqua AMSR-E measurements. H... J.. Using MODIS snow cover maps in modeling snowmelt runoff process in the eastern part of Turkey. Québec. Journal of Hydrology.. Daily snow cover estimation from Advanced Very High Resolution Radiometer Polar Pathfinder data over Northern Hemisphere land surfaces during 1982–2004. Wu. 111. Upper Marlboro. C.... L.Y. (1991). M. Cline. (2003)... Proceedings of the American Meteorological Society Annual Meeting. H. Redmond. (2006).. Y. V. MODIS/Terra Aqua Snow Cover Daily L3 Global 500 m SIN GRID V005. Model evaluation on the basis of snow cover patterns. W. Akyürekb. E. R. Remote Sensing of Environment... T. 1.. A. R.. A. Journal of Hydrometeorology.S. K. K. & Boegh. USA and northern Xinjiang. Czajkowski.. Wang.. (2003). Blöschl. T. Riggs. passive microwave. P. D. G. M. J.. Hall. Development and assessment of combined Terra and Aqua snow cover products in Colorado Plateau. Sirguey. Remote Sensing of Environment.. A. X. A. Boulder. T.. A. P... K. Nordic Hydrology. & B'ardossy. O. Validation of the MODIS snow product and cloud mask using student and NWS cooperative station observations in the Lower Great Lakes Region. 21. Spatio-temporal combination of MODIS images— Potential for snow cover mapping. Statistical evaluation of remotely sensed snow-cover products with constraints from streamflow and SNOTEL measurements.1029/2007GL029262 6 pp. J. A. X. 44. 23(3). Takeshima.. X.. & Huang. 1497−1513... D. Remote Sensing of Environment. T. R. Gao et al. Remote Sensing of Environment. H. 385.. J. G. M. D. Updated daily. X.. 3171−3179. doi:10. 21–24 January 2008. C. Validation of MODIS snow cover images over Austria. Bitner. Akyürek.. E.. Comparison and validation of MODIS standard and new combination of Terra and Aqua snow cover products in Northern Xinjiang.. E. Xie. et al. (2007).. L. T. P. (2003). Liang. (2007). D.. Zhao. B. 28 May–1 June 2007.. Hydrology and Earth System Science. 106.. & Kachi. R. W03406. Li. R.. Remote Sensing of Environment. passive microwave and scatterometer global snow products. Gafurov. Lu. Parajka. A.. & Park.. Foster. Sezai. Maryland.. G. Guo. C. G.. Hydrological Processes. Wang. Remote Sensing of Environment.. Xie. D.. 24.. A prototype AMSR-E global snow area and snow depth algorithm. (2008). (2007). Sasaki.. et al.. 1089−1104. (2006). & Tekeli. 104(3). China.. G. D. Parajka. 3750−3761. M. Water Resource Research. P. Y. Huang. IEEE Transactions on Geoscience and Remote Sensing. 64−73. Proceedings of the 60th Eastern Snow Conference. J. T. Validation of daily MODIS snow cover maps of the Upper Rio Grande River Basin for the 2000–2001 snow year.... J. M... 17. & Fernandes. China. 1866−1880.. Y. 41(2). K. Riggs. G. & Barnett. (2007). Zhang.. A. Validation of VEGETATION. N. J. 192−200.. N. M. Cloud removal methodology from MODIS snow cover product. & Liang. D.. 39(11). G. X. . Huang. 25. V. E. T.. Rhoads. 112. H. J. G. G. D. D. X. 94. H. Liu. G. & Foster. (2010). & Hendrickx. V. Colorado. MODIS. V. T. P. HDF-EOS to GeoTiff conversion tool (HEG) standalone user's guide.. G. & Kumar. K. C. Xie..8r5. W. & Salomonson.. & Salomonson. 13. 205−212). September to August 2005–2008. & Foster. Sensoyc. D. P. 11(4). A. E.3265996. S. Zhang. et al.. Tong. The Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E). Z. G. et al.. Ito. Kim.Geophysical Research Letters. K. Development and evaluation of a cloud-gap-filled MODIS daily snow-cover product. C. 341−353. (2004). D'ery.2004. Derksen. (2002). 2003 (pp. (2008). & Blöschl.. P. Chang. Automatic mapping of snow cover depletion curves using optical remote sensing data under conditions of frequent cloud cover and temporary snow. Release 3.. 1534−1547. Z. Toward improved daily snow cover mapping with advanced combination of MODIS and AMSR-E measurements.. (2007). Pu. NASDA's contribution to the EOS for global energy and water cycle studies.. Coss. Journal of Applied Remote Sensing. 3. R. L. L. 13(7).2a. Remote Sensing of Environment. Journal of Geophysical Research. K. E. Remote Sensing of Environment. (2007). 3723−3733. D. Xu. A. & Wang. & Andreassen. J. Fernandes.. Chang. et al.. S. Y.. P. G.. Hydrology and Earth System Sciences. Spencer. 1353−1360. Yang. and Cohen. 10. Z. 112.. 1675 Lobl. P. L06706.. J. 61(3–4). A. SPIE 3rd Int'l Asia-Pacific Environmental Remote Sensing Symposium. Gutman. A. Preliminary evaluation of the AFWA-NASA blended snow cover P product over the Lower Great Lakes region. Sherbrooke. Hydrological Processes. 16. J. South Dakota School of Mines and Technology. (2007). & Liang. H. K.. T. Y. 112(10). G. Zhu. Kelly. M.1. Xie. Nemani. & Blöschl. (2000).1002/hyp. L. (2004). John's. E..1117/1. & Gutknecht. Rodell. Evaluation of MODIS snow cover and cloud mask and its application in Northern Xinjiang.. A. D. Liu. doi: 10. K. & Chevallier. H. T. (2008). Remote Sensing of Environment. Soegaard. Assessment of spring snow cover duration variability over northern Canada from satellite datasets. Hubbard. An application of MODIS data to snow cover monitoring in a pastoral area: A case study in Northern Xinjiang. 216−230. Evaluation of the snow covered area data product from MODIS. June.. J. 27(12). Z. A. Maurer. IEEE Transactions on Geoscience and Remote Sensing. China. Sorman. K. (2007). S. Newfoundland. Commentary on comparison of MODIS snow cover and albedo products with ground observations over the mountainous terrain of Turkey. Dubayah. & Sormana. A.. Hydrological Processes. Boulder. Sorman. LA.. R. A. A. snowfall: The challenges. 4–6. 22(16). Wang. and scatterometer satellite data. 23−35. Liang. (2010).. L. A. China.. G. Hydrology and Earth System Sciences. A.. Y. Hangzhou. Automated monitoring of snow cover over North America with multispectral satellite data. X. M. H. 34. R. doi: 10. Riggs. K.. Romanov. X. technical paper. D. 97(2).. H.. Proceedings of the 64th Eastern Snow Conference. Kelly. Palecki. Yao.. A. Colorado USA: National Snow and Ice Data Center Digital media. Kelly. Lopez. W. Riggs. 18. 41(2). R. A. & Liang. (2009). China. (2008). R.. C. Struble. Imaoka. (2003).. X. L. 496−503. W. 114.1029/2007WR006204. A. C. Q. G. R. (2003). Simic. H.. & Czajkowski. A. E. USA: National Snow and Ice Data Center Digital media. MODIS Reprojection Tool (MRT). A.6891. Hall. E. 371(1–4). T. Advances Water Resource. (2002). USGS EROS Data Center 58 pp.. B. 514−1526.. Houborg. (2005). Accuracy assessment of the MODIS snow products. Blöschl. V. 419−429. D. R... (2009). T. (2007). T.. / Remote Sensing of Environment 114 (2010) 1662–1675 Ault. A. Shabanov. (2004). Reduction of cloud obscuration in the MODIS snow data product. A. Imaoka. Canada..1016/j. W. Zhou. Hall.Ü. 39−58. doi: 10. A. Journal of Hydrology. Pouyaud. Tekeli.. T. I.rse... Foster. MODIS/Terra observed seasonal variations of snow cover over the Tibetan Plateau. R. X.. G. J.. Hall... D.. D.. (2002). 2930−2942. doi:10. Gao. Analysis of leaf area index products from combination of MODIS Terra and Aqua data. Ishido. A. 367−381. R. Canada. 86.007. P. Journal of Atmosphere and Oceanic Technology. E.10. & Romanov. Leung. Foster.. Advances in the use of observed spatial patterns of catchment hydrological response. Liang. Riggs. Brown. T. Arnaud. Li. 5(6).... Riggs. & Hall. D. T. V 2. 230−242.. Hall. L. Global climate monitoring with the Advanced Microwave Scanning Radiometer (AMSR and AMSR-E). Pei. Taaheri. (2006).. 319−326. Xie. Kunkel. 103−116. Tedesco.. X. 297−312. Eylander. & Csiszar. Dickinson. Raytheon Company. Newfoundland.. 1064−1075. A. C. Eylander. Remote Sensing of Environment.

Sign up to vote on this title
UsefulNot useful