# 3.4 Cryosphere-related Algorithms

Size: px
Start display at page:

Transcription

1 3.4 Cryosphere-related Algorithms GLI Algorithm Description

2 3.4.1 CTSK1 A. Algorithm Outline (1) Algorithm Code: CTSK1 (2) Product Code: CLFLG_p (3) PI Name: Dr. Knut Stamnes (4) Overview of Algorithm This algorithm theoretical basis document (ATBD) describes the algorithm development for the cloudy/clear discriminator (ADEOS-II/GLI/CTSK1a) and the snow/sea-ice discriminator (ADEOS-II/GLI/CTSK1b) to be applied in the polar regions and in snow-covered mid-latitude areas. We present the theoretical basis for the algorithm, processing flow charts, and its application to MODIS Airborne Simulator (MAS) data. The calibrated radiance data at level 1B of GLI measurements in channels 8, 13, 17, 19, 24, 27, 30, 31, 34, 35, and 36 will be used as input to this algorithm. The algorithm is based on pre-defined thresholds. The snow/sea-ice discriminator works only during daytime. Fortunately, bare sea-ice conditions are expected to occur primarily in summer, and the bright polar summer will offer the opportunity to use the snow/sea-ice discriminator. The output of the cloudy/clear and snow/sea-ice discriminator algorithm will be an 8-bit word for each field of view. It includes information about whether a view of the surface is obstructed by cloud and the surface type for each pixel. The output of the algorithm is not a simple yes/no judgment for cloudy/clear discrimination. There are four levels of confidence to indicate whether a pixel is judged to be cloudy or clear. In addition, this algorithm also allows for on-screen display of a color image of the output file of a scene remotely-sensed by satellite. This color image will be a plot of the distribution of cloud and various types of surfaces. The output file of this algorithm is also an important and necessary input for cloud and surface property retrieval algorithms as well as other related studies. GLI Algorithm Description

3 B. Theoretical Description (1) Methodology and Logic Flow 1. Input and Output 1.1 Input As mentioned above, the calibrated radiance data of GLI at level 1B in channels 8, 13, 17, 19, 24, 27, 30, 31, 34, 35, and 36 will be used as input for the algorithm. If there are any incomplete and bad data in these channels for a pixel being remotely-sensed, the first bit (bit 0) of the 8-bit output word (see below) will be set to 0 for this pixel, and then the algorithm will not be executed. Thus a hole will be created in the final output map. Additionally, several ancillary data sets are required as input to the algorithm: Sun position and sensor position, i.e., zenith and azimuthal angles for both the sun and the satellite sensor The latitudes and longitudes for the current scan area A land/water map with 1 km resolution GLI surface-identification maps for the previous 8 or 4 days that will be used for the temporal uniformity test 1.2 Output The output of the cloudy/clear and snow/sea-ice discriminator algorithm will be an 8-bit word for each field of view. The output bit structure is shown in Table 2.1. The output file consists of the results of cloud mask (i.e., cloudy or clear) and the results of surface identification for cloud-free pixels. In some cases, however, it is still ambiguous to define whether a pixel is cloudy or not. For example, a pixel may be partially cloudy, or a pixel may appear cloudy at one spectral channel but appear cloud-free at another channel. A brightness temperature difference test, based on the 10.8 and 3.7 µm channels, is sensitive to water clouds, but not to thin cirrus. In contrast, the 1.38 µm channel is sensitive to the presence of high level cirrus due to the strong water vapor absorption. Thus, multi-spectral approaches are expected to be particularly useful for cloud detection in the polar regions. The output of the algorithm is not a simple yes/no judgment for cloudy/clear discrimination. There are 4 levels of confidence to indicate whether a pixel is judged to be cloudy or not (bit 2-1 of the output word) based on the analysis of the results from different spectral tests. A brief description of the output bits is given as follows (see Table 2.1): GLI Algorithm Description

4 Table 2.1 The output of the cloudy/clear and snow/sea-ice discriminator algorithm. The output is an 8-bit word, i.e., Bit No. Bit output Meaning Remark 0 not executed bit 0 1 executed 00 clear sky 01 high-confidence cloudy confidence = 100% bit middle-confidence cloudy 50% confidence <100% 11 low-confidence cloudy 0% < confidence < 50% 000 snow over ice 001 sea ice bit cloud shadow on snow 011 land with vegetation 100 open ocean 101 snow over land 110 snow over land contaminated with vegetation 111 bare land with less vegetation bit 7-6 Spares Bit 0: If the sensor-measured radiance of a pixel is deemed to be bad, bit 0 will be set to 0, and the algorithm will not be executed for this pixel. If the algorithm is executed normally bit 0 will be set to 1. Bit 2-1: Bit 2-1 represents the final results of the cloudy/clear discrimination processing. For a given pixel bit 2-1 indicates clear or cloudy with different confidence levels (see Table 2.1). Based on the threshold value for each individual cloud test, a confidence flag is assigned. There is a linear interpolation between the high-confidence-cloud threshold (100%) and the clear-sky threshold (0%). If the confidence is 100%, bit 2-1 is set to 01. If the confidence is between 50% and 100%, bit 2-1 is set to 10. If the confidence is between 0% and 50%, bit 2-1 is set to 11. For the case with confidence of 0%, bit 2-1 is set to 00, and the pixel is assumed to be clear (cloud-free). Because various test techniques (or methods) may be sensitive to different kinds of cloud, the final determination will be made by considering the maximum confidence of all tests applied. This implies that the algorithm will be cloud conservative. For a pixel with low- GLI Algorithm Description

5 confidence cloud, additional spatial and temporal consistency tests will be implemented for improving the cloudy/clear discrimination capability. Bit 5-3: For a clear pixel (bit 2-1 is set to 00), the algorithm will assign the surface type. Generally, the thresholds for reflectance and brightness temperature can be used to distinguish pixels with snow/sea-ice from pixels with the land/open-ocean. The land/water map will be used to flag land and ocean. Further, the snow/sea-ice discriminator will be used to distinguish the bare sea-ice from snow-covered sea-ice and/or open water in ocean areas. There are five surface types that will be tagged by this algorithm: snow (including snow-covered sea-ice), bare sea-ice (snow-free), land (including vegetation), open ocean, and cloud shadow over snow. The cloud shadow test will be implemented only over snow during daytime. In addition, this algorithm can also offer the ability to plot a color image of the processing results. The color image will be a plot of the distribution of cloud and various types of surface. It provides a visualization useful for testing our understanding of the algorithm. The output file of this algorithm is also an important and necessary input for cloud retrieval and surface property retrieval algorithms as well as other related studies. 2. Implementation of the Algorithm The algorithm includes four parts: cloudy/clear discriminator, snow/sea-ice discriminator, cloud shadow detection, and surface classification (as shown in Figure 2.1). It can be used to identify clear scenes and to provide surface classification. Sea-ice will be classified into bare sea-ice and snow-covered sea-ice. This information will be saved in an output file consisting of an 8-bit word for each pixel that is required for retrieval of cloud and surface properties. The cloudy/clear discriminator consists of a daytime mode and a nighttime mode. In each mode, there are two independent tests for detection of low cloud and thin cirrus cloud. Because surface temperature inversions often exist under clear-sky conditions during polar winter, a temperature inversion test will be included in both the daytime mode and the night time mode to improve the capability of cloud detection. The result of each test is expressed as a confidence level that indicates the likelihood that the pixel is contaminated with cloud. The final confidence level is determined by the maximum confidence level resulting from applying several tests. Most pixels have either confidence 1 (high confidence cloud) or 0 (clear pixel). Pixels with an intermediate confidence value, tend to be found at cloud boundaries. For clear pixels, surface classification will be applied to tag the surface as land, open ocean, snow-covered land, or snow/ice-covered ocean. Each clear pixel will be classified as being GLI Algorithm Description

6 land or ocean according to its location in the land/water map. Simple reflectance and/or brightness temperature threshold tests are used to distinguish snow/ice-covered surface from land or open ocean. Figure 2.1 The flow chart of algorithm for single pixel test. GLI Algorithm Description

7 Cloud shadow detection is applied for low albedo surfaces. Further work on the cloud shadow detection over snow/ice will be needed in the future. The snow/sea-ice discriminator relies on visible and near infrared channels to distinguish bare sea ice from snow-covered sea ice. The NDII technique and the visible reflectance ratio test used in the algorithm will only be applied during daytime. Bare sea-ice appears mainly during the summer. Thus, the long daylight hours of the polar summer is an advantage since the snow/sea-ice discriminator is limited to daytime use. An outline of the algorithm has been shown in Figure 2.1. The implementation steps of algorithm are as follows: I. Input GLI calibrated radiance data at level-1b and other ancillary data, including A. solar and sensor position, and the location of the scan area B. land/water map C. GLI surface identification maps for the previous 8 or 4 days II. Apply cloudy/clear discriminator to set confidence flag as cloudy or clear. A. determine whether the pixel is in a day or night regime B. apply single cloudy/clear test to assign confidence level for each test, depending on its value relative to the threshold C. determine final confidence flag, which is the maximum confidence of all applied tests III. Apply surface classification algorithm to classify surface into high albedo surface (snow and ice) and low albedo surface (bare land, land with vegetation and open ocean). IV. Apply cloud shadow test to tag shaded pixels. V. Apply snow/sea-ice discriminator to tag bare sea-ice and snow-covered sea-ice. VI. Output the results of the algorithm, which is an 8-bit word for each pixel. The algorithm has been applied to MAS (MODIS Airborne Simulator, deployed on-board the ER-2 aircraft) data (ARMCAS, April-95 Alaska, WINCE). MAS has a spatial resolution of approximately 50 m, which is much higher than the ground resolution (1 km) of the GLI sensor. Results from these applications should aid in algorithm development. A comparison of NDSI technique and brightness temperature difference test of 10.8 and 3.7 µm has been presented to verify the feasibility of using brightness temperature difference test instead of NDSI technique. The algorithm has been applied to high latitude areas (ARMCAS and April- GLI Algorithm Description

8 95 Alaska data) as well as to mid-latitude areas (WINCE data). It is feasible to extend this algorithm to snow-covered mid-latitude areas. Due to the absence of winter data and the 6.7 µm channel in MAS, the temperature inversion test could not be tested by use of MAS data. (2) Physical and Mathematical Aspects of Algorithm 1. Objective There are two main objectives of this investigation: (i) to determine whether a given FOV (field of view) is obstructed by clouds, and (ii) for cloud-free pixels to distinguish the bare seaice from snow-covered sea-ice. The algorithm yields an 8-bit word output for each pixel to describe the sky condition (i.e., whether the pixel is obstructed by clouds) and the surface type for a given pixel under cloud-free conditions. The surface will be classified into five possible types: snow, sea-ice, cloud shadow, land (tundra), and open ocean. When sea-ice is covered by snow (even only a few centimeters), the surface radiative characteristics will be similar to snow (Jin et al., 1994). Thus, sea-ice covered by snow will be classified as a snow surface, while only bare sea-ice is classified as sea-ice. 2. Background Our development of the cloudy/clear discriminator and the snow/sea-ice discriminator for ADEOS-II/GLI has benefited from previous studies in this field (Hall et al., 1996; Ackerman et al., 1997). Since some of the GLI channels are similar to those of MODIS (The Moderate Resolution Imaging Spectrometer), we will adopt the results of previous studies in our work. Compared to AVHRR (Advanced Very High Resolution Radiometer) that has only five channels, the multiple GLI channels will allow us to extract more information about the cloud and snow/sea-ice surface in the polar regions which has been considered as a difficult remote sensing problem. However, many methods and techniques as well as related results concerning a cloudy/clear discriminator developed for the AVHRR can be employed and modified for use with the GLI. The ISCCP (International Satellite Cloud Climatology Project) cloud masking algorithm, as described by Rossow (1989) and Rossow and Garder (1993), relies on comparing the observed radiances with their corresponding clear-sky values in two channels (0.6 µm and 11 µm). Rossow and Garder (1993) summarized the five steps employed for ISCCP cloud detection as follows: (i) space contrast test on a single infrared image; (ii) time contrast test on three consecutive visible images; (iii) accumulation of space/time statistics for infrared and visible images; (iv) construction of clear-sky composites for the infrared and visible every five days at each diurnal phase and location; and (v) radiance threshold for the infrared and visible for each pixel. GLI Algorithm Description

9 The AVHRR Processing scheme Over cloud Land and Ocean (APOLLO) (Kriebel and Saunders, 1988; Kriebel et al., 1989; Gesell, 1989) uses AVHRR channels 1 through 5 based on five threshold tests, i.e., a brightness temperature threshold test, the reflectance ratio of channels 2 to 1, the temperature difference between channels 4 and 5, and the spatial uniformity over ocean (Kriebel et al., 1989). A pixel is defined as cloud contaminated if it fails in any single test, so in that sense the APOLLO scheme is cloud conservative. The NOAA CLAVR algorithm (Phase I) relies on multispectral decision, channel differences, and spatial differences, and then employs a series of sequential decision tree tests with all five channels of the AVHRR (Stowe et al., 1994). A 2 x 2 global area coverage (GAC) pixel array is used for identification in such a way, that if all four pixels satisfy just one of the cloud tests, then the array is labeled as 100% cloudy or restored-clear, otherwise, it is labeled as mixed or 50% cloudy. Subsequent phases of CLAVR will use dynamic thresholds and pixel by pixel classification. A cloud-contaminated pixel is radiatively typed as belonging to low stratus, thin cirrus or a deep convective cloud system. Several algorithms have also been developed for cloud clearing for the TIROS-N Operational Vertical Sounder (TOVS). For example, the fifth version of the International TOVS Processing Package (ITPP-5) uses collected AVHRR and HIRS/2 (High-Resolution Infrared Sounder) to cloud-clear the HIRS/2 footprints (3 x 3 retrieval box). Several methods, such as a brightness temperature (BT) or an albedo threshold test, a BT difference test or a BT, a skin temperature difference test, and a difference test employing BT derived from AVHRR and HIRS/2, are used to determine cloudy sky condition. The nature of the cloud cover is determined in further steps. Operational GOES products by NESDIS use an array of n x n continuous pixels and use a method similar to that of ITPP-5 (Ackerman et al., 1997). As the high albedo of snow cover presents a good contrast with most other natural surfaces, snow cover was observed in the first image obtained from the TIROS-1 weather satellite following its launch in April 1960 (Singer and Popham, 1963), and has been mapped in the Northern Hemisphere on a weekly basis since 1966 by NOAA. A variety of sensors, including AVHRR and passive-microwave sensors, have been widely used for the detection of snow cover. Various techniques, ranging from visual interpretation, multispectral image classification, decision trees, change detection, and ratios (Kyle et al., 1978; Bunting and d Entremont, 1982) have been developed to map snow cover with remotely sensed data. Spectral-mixture modeling is also used for subpixel classification of snow in a scene (Nolin et al., 1993; Rosenthal, 1993). The snow-mapping algorithm (SNOMAP) developed for MODIS relies on use of an NDSI (Normalized Difference Snow Index) test and a decision rule to identify snow in each 500 m MODIS pixel (Hall et al., 1995). The use of the NDSI for snow/cloud discrimination is based GLI Algorithm Description

10 on the fact that the snow reflects visible radiation more strongly than it reflects radiation in the middle-infrared part of the spectrum, while cloud reflectance remains high in these two spectral regions. Thus, a low value of the NDSI is indicative of cloud, and a relatively higher value of snow cover. However, a visible threshold test must be conducted to separate snow cover from land/open ocean. The NDSI test is of limited value over dense forest cover, or in the presence of thin cirrus clouds. Other advanced classification techniques (Nolin et al., 1993) permit subpixel classification and improve the identification of fractional snow cover. It is likely that spectral-mixture modeling or another advanced classification technique that does not rely on thresholding could be implemented after launch of MODIS (Hall et al., 1995). Some of the above cloud masking algorithms have been incorporated into current global cloud climatologies and have been run in an operational mode over long time periods. We have also benefited from other cloud detection studies, e.g., the algorithms developed for MODIS, in our development of a cloudy/clear discriminator for the GLI. However, due to the special difficulty associated with cloud detection over high albedo surfaces in the polar regions and the characteristics of the GLI multispectral information, some improvements to the previous algorithms have been made and used with the specific GLI channels. Since most of our data processing studies are based on MAS data and radiative transfer simulations, further improvements may be possible when data from the GLI become available. The detection of cloud shadows is a difficult problem in the development of a proper cloudy/clear discriminator. Theoretically, they can be computed from the given view angle, solar angle, distribution of cloud edges, and cloud altitude, but it takes considerable computing time. There are two thresholds used for cloud shadow detection over land, which are based on the reflectance at channels 13 (0.678 µm), 19 (0.865 µm), and 24 (1.05 µm): R /R > 0.9; R 1.05 < 0.12 (Ackerman et al., 1997). Cloud shadow detection over snow is currently at an immature stage and will require substantial development beyond the scope of this project. 3. GLI Characteristics The GLI has 36 channels with channels 1 29 in the visible and near-infrared (NIR), and the rest of the channels in the thermal-infrared ranging from 3 to 12 µm. The specifications of the GLI channels used here are shown in Table 2.2. There are some channels that are similar among GLI, MODIS and AVHRR (see Table 2.2). Due to this similarity, our work is benefiting from previous experience and results by others in algorithm development. GLI Algorithm Description

11 Table 2.2 GLI channels to be used in the algorithm GLI channel λ (µm) Δλ (µm) GR (km) Purpose surface classification, cloudy/clear and snow/sea-ice discriminator surface classification, cloud shadow, cloudy/clear and snow/sea-ice discriminator cloudy/clear discriminator snow/sea-ice discriminator and cloud shadow cloud shadow, snow/sea-ice discriminator cloudy/clear discriminator cloudy/clear discriminator cloudy/clear discriminator cloudy/clear discriminator cloudy/clear discriminator cloudy/clear discriminator MODIS channel AVHRR channel MAS channel (1.88 µm) Figure 2.2 shows simulated planetary albedoes for various types of surfaces and water/ice clouds in the µm spectral range. The location and bandwidth of GLI channels 1-29 are also shown in Figure 2.2. The 36 GLI channels offer the opportunity to use multispectral approaches for cloud detection. The analysis of cloud cover in the polar regions from satellite data is more difficult than at lower latitudes because the visible and thermal contrasts between the cloud cover and the underlying surface are frequently quite small. However, a notable difference between cloud and snow can be found at 1.6 µm. Based on this information, the NDSI technique could be used to distinguish clouds from snow surfaces (Hall et al., 1995). Unfortunately, the ground resolution of GLI channel 28 (1.64 µm) has been designed as 250 m, GLI Algorithm Description

12 rather than 1 km which is the ground resolution for all channels we used here. So we cannot use the NDSI technique in this algorithm. Another brightness temperature difference test using channel 30 (3.7 µm) and channel 35 (10.8 µm) will be adopted for the cloudy/clear Figure 2.2 The planetary albedoes of various types of surfaces and clouds in the spectral range between 0.3 and 2.5 µm for solar zenith angle θ=60. The location and bandwidth of GLI channels 1-29 are also shown at the top part of the figure with black dots or lines. discriminator instead of the NDSI technique. The cloudy/clear discrimination can be accomplished using GLI channels 8, 13, 19, 24, 27, 30, 31, 34, 35 and 36. These channels are employed for detecting low cloud, high cloud, and cloud shadow. The principal differences in planetary albedo between snow and ice occur at NIR bands ( µm), as shown in Figure 2.2. As a result of the strong absorption of sea ice in the NIR, the reflectance over ice is much lower than over snow. This difference provides the theoretical basis for the snow/sea-ice discriminator. Based on the different albedo between snow and seaice, GLI channels 8, 13, 19 and 24 are used for snow/sea-ice discrimination. We use a new index called the NDII (Normalized Difference Ice Index) as our main test for snow/sea-ice discrimination. The NDII test will be used only during day time. Fortunately, sea-ice is covered by snow most of the year. Bare sea-ice appears primarily in the summer season. The GLI Algorithm Description

13 bright polar summer offers a possibility to apply the NDII-based snow/sea-ice discriminator which is effective only during day time. GLI channels with 1 km ground resolution will be used for this algorithm. GLI will provide measurements at the same position every four days. The four-day recovery schedule is also used to test for temporal uniformity for the confirmation of surface type. MAS data are used for the validation of the algorithm. The MAS data set used in this study includes ARMCAS (Arctic Radiation Measurement in Column Atmosphere-surface System), ALASKA_April_95 (i.e., snow, glaciers and sea-ice measurements, March-April 1995, in north central Alaska), and WINCE (WINter Cloud Experiment, Jan. 23 to Feb. 13, 1997, over the Great Lakes area) data. These data sets provide the opportunity to apply and test the cloud detection and surface identification algorithm over polar and mid-latitude snow-covered areas. 4. Physical aspects of the algorithm This algorithm is different from the MODIS cloud-mask algorithm, which is based on a given surface condition. We have designed an algorithm which is independent of surface conditions so that is will be applicable to the variable surface conditions encountered in the polar regions during the spring, summer and autumn. We also employ multispectral ratios and threshold-based methods to obtain a simple and effective algorithm. The algorithm consists mainly of the cloudy/clear discriminator (CTSK1a), the snow/seaice discriminator (CTSK1b), and a surface classification. The last component goes beyond the tasks CTSK1a and CTSK1b that we committed to do as part of this project. We offer this simple surface classification algorithm to help identify high-albedo ocean areas from the clear pixels. The snow/sea-ice discriminator algorithm will be implemented only for these highalbedo ocean areas, to distinguish the bare sea-ice from snow-covered sea-ice. A flow chart of the algorithm for a single pixel test is shown in Figure 2.1. The algorithm will start with singlepixel (1 km FOV) tests. For pixels with low-confidence cloud, a spatial coherence test may be applied on 3 x 3 pixel segments to improve the confidence of the cloud detection. For nomenclature, we shall denote the satellite-measured radiance as I and the normalized reflectance as R, and we shall refer to the infrared radiance as brightness temperature (equivalent blackbody temperature determined using the Planck function) denoted as BT. Subscripts refer to the wavelength in µm at which the measurement is made. We define the normalized reflectance as: πi R λ = λ S λ cosθ 0 GLI Algorithm Description

14 where S is the extraterrestrial solar irradiance. A threshold will be set for each single test based on the normalized reflectance R when the channel wavelength is less than 3 µm, and based on brightness temperature when the channel wavelength is larger than 3 µm. 4.1 The Cloudy/Clear Discriminator The detection of clouds in the polar regions or in snow-covered mid-latitude areas is difficult because of insufficient albedo contrast between the cloud and the underlying snow or sea ice surface. Since snow and ice are considerably more reflective in the visible than in the NIR part of the spectrum and the reflectance of clouds remains high in the NIR, the Normalized Difference Snow Index (NDSI) technique based on a channel at 1.64 µm is used to distinguish snow from clouds (Hall et al., 1995). Due to the limitation of ground resolution of GLI channel 28 (1.64 µm), we cannot use the NDSI technique in our cloudy/clear discriminator. A brightness temperature difference (BT 10.8 and BT 3.7 ), which was used only in night mode in our earlier version, will be used also during daytime, instead of the NDSI technique. A comparison between the NDSI technique and the brightness temperature difference, based on MAS data, has demonstrated the usefulness of the brightness temperature difference test. Additionally, we use the 1.38 µm channel for thin cirrus detection during daytime and we use four thermal channels (3.7, 8.6, 11, and 12 µm) for cloud detection during night. Under clear-sky conditions, strong surface temperature inversions often exist in the polar regions during winter. Clouds generally tend to inhibit the formation of the inversion. Therefore, the presence of a strong, low-level temperature inversion is a good indication of a clear sky condition. The 6.7 and 10.8 µm channels will be used for temperature-inversion detection in polar winter for both the daytime and the nighttime modes. A detailed analysis of the temperature-inversion test will be deferred (see below). A combination of these tests constitutes the main part of the algorithm, since none of them can be considered as a perfect technique for detecting all kinds of clouds under different environmental conditions. For example, some are more reliable for low-cloud detection, while others are better for high-cloud detection. The result of each single test is expressed as a confidence level ranging from 1 (cloudy pixel with high confidence) to 0 (clear pixel). The final confidence flag will be assigned by considering the maximum confidence of all tests applied. This is called the cloud conservative approach. A common result of the cloudy/clear discrimination from this algorithm is either highconfidence cloudy or clear, except for cases in which there are some low-confidence clouds at cloud boundaries, or thin, low, or small clouds exist. In these cases, spatial and/or temporal continuity tests need to be conducted. The cloudy/clear discrimination algorithm has a daytime GLI Algorithm Description

15 mode and a nighttime mode depending on the solar zenith angle. Daytime is defined by limiting the solar zenith angle to θ 0 < 85. The daytime mode includes brightness temperature difference (using the 3.7 and the 10.8 µm channels) and the 1.38 µm channel thin cirrus tests; and the nighttime mode includes infrared brightness temperature difference tests, using channels at 3.7 and 10.8 µm. The temperature-inversion test is also included in both the daytime and the nighttime modes. Figure 2.3 shows a flow chart of the cloudy/clear discriminator algorithm. Figure 2.3 The flow chart of cloudy/clear discriminator. F1, F2, F3, and F are the confidence levels for individual test and final result, respectively. GLI Algorithm Description

16 4.1.1 Brightness Temperature Difference Test (BT 10.8 and BT 3.7 ) At high latitudes, infrared window tests based on satellite-measured radiances are difficult due to the cold surface, and especially the small temperature contrast between cloud and snow surface in winter. However, a nighttime polar (Antarctic) cloud/surface discrimination algorithm has already been applied by Yamanouchi et al. (1987). This algorithm is based upon the brightness temperature difference between the AVHRR 3.7 and 11 µm channels and between the AVHRR 11 and 12 µm channels. GLI channel 30 (3.7 µm) measures radiances in the window region near µm, so we can use GLI channels 30 and 35 (10.8 µm), another window region, to detect the presence of clouds. In particular we can detect partial cloudiness or thin cloud cover at night based on the difference between BT 10.8 and BT 3.7. The MAS data at 3.7 µm (channel 31) and 11.0 µm (channel 45) will be used for validation and determination of the threshold. Another method for determining the threshold is based on modeling results. Small or negative differences BT 10.8 BT 3.7 are observed only for cases when an opaque scene (such as thick cloud or the surface) fills the field of view of the sensor. Negative difference BT 10.8 BT 3.7 occurs at night over extended clouds due to the lower cloud emissivity at 3.7 µm. Large negative difference BT 10.8 BT 3.7 occurs during daylight hours because of the large reflection of solar radiance at 3.7 µm. This test will be used in both daytime and nighttime modes. Because of the reflection of solar energy at 3.7 µm during daytime, the thresholds used for daytime and nighttime modes are quite different Thin Cirrus Test (1.38 µm) channel The detection of thin cirrus is a challenging problem, especially in polar regions. Fortunately, GLI channel 27 (1.38 µm) provides an opportunity to detect the presence of thin cirrus in the upper troposphere under daytime viewing conditions by use of reflectance thresholds (Gao et al., 1993). The premise of this test comes from the strong water vapor absorption in the 1.38 µm region. Therefore, most of the earth's surface is obscured at this channel. With relatively little of the atmosphere s moisture located high in the troposphere, high clouds appear bright. Simple low and high reflectance threshold will be used to separate thin cirrus from clear scenes as well as scenes with thick cloud. The MAS 1.88 µm channel will be used in conjunction with theoretical simulations for validation and for determining the threshold, because strong water vapor absorption exists in both the 1.38 µm channel and the 1.88 µm channel. GLI Algorithm Description

17 4.1.3 Brightness Temperature Difference Test (BT 8.6 BT 12 ) The 1.38 µm channel thin cirrus test can not be used for cloudy/clear discrimination during the dark winter time in the polar regions. Instead we have to use infrared channels for this purpose. Because the sensitivity of the refractive index for ice and water vapor is different in the window region, their absorption peaks are in opposite halves of the window region around 8-12 µm. The difference in brightness temperature between GLI channels 34 (8.6 µm) and 36 (12 µm), BT 8.6 BT 12, can therefore be applied during night time to discriminate between high level clouds and cloud-free conditions Temperature Inversion Test (BT 6.7 and BT 6.7 BT 10.8 ) Normally, there is strong temperature inversion under clear-sky conditions in polar winter (Curry et al., 1996). Thus, a temperature-inversion test is another way to achieve cloudy/clear discrimination. Infrared channels whose weighting functions peak low in the atmosphere will often have a larger BT than a window channel (e.g., BT 6.7 BT 10.8 ). In clear sky situations, the 6.7 µm radiation measured by the satellite instrument is emitted by water vapor in an atmospheric layer located approximately between hpa (Soden and Bretherton 1993; Wu et al., 1993), while 6.7 µm radiation emitted by the surface or low clouds is absorbed in the atmosphere and is not sensed by the satellite instruments. On the contrary, the measured 10.8 µm radiation under clear sky originates primarily at the surface. Strong inversions in polar winter under clear-sky conditions lead to large negative differences in BT 10.8 BT 6.7 (below - 10 K) that can be used to detect cloud-free conditions. The GLI channels 31 (6.7 µm) and 35 (10.8 µm) will be used for this test Confidence Flags Almost all tests mentioned above rely on the pre-defined thresholds. Instead of a single threshold, the user can obtain information about how much confidence can be placed in the results. The confidence of every single test is assigned a value between 0 and 1, representing increasing confidence that clouds are present. Figure 2.4 is a graphical representation of how a confidence level is assigned for a test. As shown in Figure 2.4, an observational value greater than γ corresponds to high confidence that a cloud is present and is assigned a confidence level of 1. An observed value less than α is clear and will be assigned a confidence level of 0. These high-confidence cloudy and clear thresholds, γ and α respectively, are determined from observations and/or theoretical simulations. Observational values between α and γ are assigned a confidence level between 0 and 1. Assignment is based on a linear interpolation. Each of the tests mentioned above returns a confidence level ranging from 1 (high confidence cloud) to 0 (clear-sky). Based on the cloud conservative approach, we set final GLI Algorithm Description

18 confidence level as the maximum confidence level of all applied tests (shown in Figure 2.3). The final confidence level will determine the final flag as cloudy or clear for the output file (bit 2-1). Figure 2.4 A graphical representation of confidence level. 4.2 Theoretical Description of the Snow/Sea-ice Discriminator Snow exhibits strong reflectance in the visible region and strong absorption in the NIR region. The NDSI technique has been successfully applied to distinguish snow from clouds as discussed in previous sections. Sea ice has characteristics similar to snow. It is more difficult to distinguish sea-ice from snow than detecting clouds over snow. However, there is a difference in albedo between snow and sea-ice in the region of µm (see Figure 2.1). Due to the strong absorption of sea-ice, the planetary albedo of sea-ice drops rapidly at 0.6 µm, while snow maintains relatively high albedo until 1.2 µm. Based on this characteristic, we GLI Algorithm Description

19 employ the NDII technique (Normalized Difference Ice Index) (see section 4.2.1) as the main test for snow/sea-ice discrimination. A reflectance ratio test (see section 4.2.2) is an auxiliary test used in the snow/sea-ice discriminator algorithm. The two tests in the snow/sea-ice discriminator will only be used during daytime Normalized Difference Ice Index (NDII) Test Under clear-sky conditions, the principal differences in planetary albedo between snow and sea-ice occur in the NIR region between 0.6 and 1.2 µm, as shown in Figure 2.2. Snow and sea ice have similar reflectance for wavelengths shorter than 0.6 µm. As a result of the strong seaice absorption in the NIR, the reflectance over sea-ice is much lower than over snow in that wavelength region. Thus channel 24 (1.05 µm) can be used to distinguish snow from sea ice. In addition, we also need to employ another band ( µm) to distinguish ocean from snow/sea-ice. Since ocean has lower albedo than snow/sea-ice in this visible region (see Figure 2.2), GLI channel 8 (0.545 µm) can be used for the separation of snow/sea-ice and ocean. Actually, there are plenty of melt ponds on the sea ice due to the melting during summer. The visible radiances for such cases will have values between that corresponding to the sea-ice case and that corresponding to the open-ocean (water) case. However, albedo of water is low at both channel 8 (0.545 µm) and channel 24 (1.05 µm), while albedo of sea ice is high at channel 9 and low at channel 24, and albedo of snow is high at both channel 8 and channel 24 (see Figure 2.2). We then define a snow/sea-ice discriminator using NDII that is similar to the case of NDSI: NDII = R R 1.05 R R 1.05 where R and R 1.05 are the normalized reflectances of channels 8 and 24, respectively. A reference radiance at channel 13 (0.678 µm) is required to distinguish snow from dark surfaces. For open ocean, the reference radiance is always low. When the reference radiance is high and the NDII is below a pre-defined value threshold, the current pixel is covered by snow. When the reference radiance is high and the NDII is larger than this threshold value, the pixel is covered by sea ice Reflectance Ratio Test The reflectance ratio test employs the ratio of satellite-measured radiance between channels 19 (0.865 µm) and 13 (0.678 µm) (i.e., R /R ). This test is based on the fact that the spectral reflectances at these two wavelengths are similar over snow or water (ratio is near 1) and different over sea-ice. GLI Algorithm Description

20 4.3 Surface Classification There is a simple surface classification included in the algorithm. It will be applied to all clear pixels tagged by the cloudy/clear discriminator. Visible and infrared (needed during night time) threshold tests can be applied to discriminate high albedo surfaces from low albedo surfaces. We use GLI channels 8 (0.545 µm) and 35 (11 µm) for visible and infrared threshold tests, respectively. The land and ocean area will be determined by the land/water map. Thus, the classification of four types of surface (open ocean, land, snow-covered land, and snow/seaice covered ocean) can be accomplished by using two threshold tests and the land/water map. It is necessary to tag the snow/sea-ice-covered ocean area prior to applying the snow/sea-ice discriminator. C. Practical Considerations (1) Programming Requirements Table 3.1 provides information about the software generated from this algorithm. Table 3.1 Program Requirements Program Memory 400 KBytes Program Size 100 KBytes Required Channels 8,13, 17, 19,24,27,30,31,34,35,36 Necessary/Ancillary Data land/water map Expected Disk Volume 5 MBytes Special Programs or Subroutines mask.f Thresholds list thresholds.inc (2) Calibration and Validation 1. Cloudy/clear Discriminator for MAS data For the cloudy/clear discriminator there are two modes applied during daytime and nighttime, respectively. A brightness temperature difference test using channels at 10.8 and 3.7 µm is used for both daytime and nighttime modes to distinguish cloud from snow. In addition, a 1.38 µm channel test and a test employing the brightness temperature difference between the 8.6 and the 12 µm channels are used to detect the presence of thin cirrus in both daytime and nighttime modes, respectively. Since nighttime MAS data are not available, we have used the infrared data obtained during daytime data for validation and improvement of the algorithm. Because the threshold of the daytime mode is quite different from that of the nighttime mode, MAS data cannot be used to establish the correct nighttime threshold. GLI Algorithm Description

21 1.1 Comparison between NDSI test and BT 11 BT 3.7 test Figure 3.1 shows the results from one scene of WINCE data using the NDSI test and the brightness temperature difference test. The scene used is the observation made on February 9, 1997 over Lake Erie. We use the MAS channel at 1.88 µm instead of GLI channel at 1.38 µm for the thin cirrus test because MAS does not have a channel at 1.38 µm and there is strong water vapor absorption at both the 1.38 µm and the 1.88 µm channels. High radiance at the 1.38 µm channel (1.88 µm channel of MAS) indicates the presence of high level cloud (including thin cirrus cloud). For the left panel of Figure 3.1, we used the NDSI test as well as the 1.88 µm channel test for the cloudy/clear discrimination. For the right panel of Figure 3.1, we used the test based on the brightness temperature difference between channels at 11 and 3.7 µm as well as the 1.88 µm channel test for the cloudy/clear discrimination. The different colors used in Figure 3.1 represent the results of cloudy/clear discrimination (see Table 3.2). Comparing the two panels in Figure 3.1, one finds that the difference between the two images (i.e. the two panels in Figure 3.1) is less than 10%. This indicates that the brightness temperature difference test may generally work as well as the NDSI test. Therefore, we can use the brightness temperature difference test as the main test instead of the NDSI test for cloudy/clear discrimination. Table 3.2 Description of colors in cloud mask image color red blue yellow green sea green coral magenta cyan meaning snow over ocean sea ice land high confidence cloud middle confidence cloud low confidence cloud snow over land water (open ocean) 1.2 Comparison between 1.38 µm test and BT BT 12 test In cloudy/clear discriminator, we use the 1.38 µm test for the daytime mode and the test employing the brightness temperature difference between the 8.6 and 12 µm channels for nighttime modes. These two tests offer the potential to improve the detection of thin cirrus. Here we show a comparison between these two tests for a scene of the MAS ARMCAS data. This scene obtained on June 11, 1995 over the North Slope of Alaska. Figure 3.2 shows the cloud mask resulting from use of the 1.38 µm test, the BT 8.6 BT 12 test, and the BT 11 BT 3.7 GLI Algorithm Description

22 test, respectively. The left panel of Fig. 3.2 corresponds to the 1.38 µm test for the daytime mode, the middle panel corresponds to BT8.6 BT12 test for the nighttime mode, and the right panel corresponds to BT11 BT3.7 test for both daytime mode and nighttime mode. Cirrus cloud should have a high value at 1.38 µm and BT8.6 BT12. At the top of image, there are some water clouds detected by 1.38 µm test, but not by BT8.6 BT12 test. It implies that 1.38 Figure 3.1 An example of a cloudy/clear discriminator image derived from MAS WINCE data over the Lake Erie on Feb. 9, 1995 (track 3). The left panel is for the NDSI test. The right panel is for the brightness temperature difference test. µm test is not only sensitive to cirrus cloud, but also to some thick low cloud. These water clouds have also been detected by BT11 BT3.7 test which is a main test for both daytime and nighttime mode (See right panel of Fig. 3.2). Figure 3.3 shows two images of cloud mask results using the same data as shown in Fig. 3.2 for daytime mode (left panel) and nighttime mode (right panel). From Fig.3.3, we have found that both daytime mode and nighttime mode yield similar cloud mask images. GLI Algorithm Description

23 Figure 3.2 An example of a cloudy/clear discriminator image derived from MAS ARMCAS data collected over the North Slope of Alaska on June 11, 1995 (track 4). The left panel is for the 1.38 µm test, the middle panel is for the BT 8.6 BT 12 test, and the right panel is for BT 11 BT 3.7 test. From the results of the two cases in Figs. 3.1 and 3.2, we have found that we can successfully carry out cloudy/clear discrimination using BT 11 BT 3.7 test instead of NDSI test in daytime mode. We have also found that both the daytime and the night modes of the cloudy/clear discriminator algorithm seem to work well over snow snow/ ice surfaces in the polar regions as well as in mid-latitude areas. 2. Snow/sea-ice Discriminator for MAS data Under clear-sky conditions, snow/sea-ice discrimination over ocean area will be based on the NDII test and the ratio-threshold test. Validation of this algorithm has been carried out by using MAS data obtained in the ARMCAS and Alaska-April-95 campaign. MAS channels 2 (0.653 µm) and 9 (0.945 µm) are employed in the NDII test to mimic GLI channels 8 (0.545 µm) and 24 (1.05 µm). The reflectance ratio between MAS channels 2 (0.653 µm) and 5 GLI Algorithm Description

24 Figure 3.3 An example of a cloud mask iamge derived from MAS ARMCAS data over North Slope of Alaska on June 11,1995 (track 4). The left panel is for daytime mode, and right panel is for nighttime mode. (0.781 µm) is also employed to mimic GLI channels 13 (0.678 µm) and 19 (0.865 µm), as the second test for discriminating snow from bare sea-ice. One case is presented here to indicate that the NDII test and the ratio test can be used to discriminate snow from bare sea-ice. This scene is an observation made on April 8, 1995 (track 14) over the Bering Sea (Fig. 3.4). This figure shows the ice mask combining the NDII GLI Algorithm Description

25 test and ratio test, The algorithm is sea-ice conservative. The colors in image have been described in Table 3.2. Figure 3.4 An example of snow/sea-ice discriminator derived from MAS data over the Bering Sea of Alaska on April 8, 1995 (track 14). This shows the results of the cloud mask and snow/sea-ice discriminator. Prior to generating the sea ice mask, pixels with high confidence cloud occurrence (green area) are identified based on the BT 11 - BT 3.7 test and the 1.88 µm test. Then, a reference channel (MAS channel 2, corresponding to GLI channel 13) is used to discriminate sea ice from open water. Finally, the NDII and the ratio tests are applied to mark the bare sea ice GLI Algorithm Description

26 surface (blue area). The cyan area is the open ocean, caused by ice lead or pond after the snow over sea ice melts, when the reflectance in MAS channel 2 is low. The red area is snowcovered area when the reflectance in channel 2 is high. In Fig. 3.4, St. Lawrence Island covered by snow is just at the top of panel. There is open water along the coast. Near the coast, there are many floating ice sheets. This is a scene under clear-sky condition, except for a small amount of cumulus clouds over the ocean surface. Snow-covered sea-ice was present at the bottom of the scene. From the case discussed in this section, it is obvious that the sea-ice can be discriminated from snow and water by the algorithm described above. (3) Quality Control and Diagnostic Information (4) Exception Handling (5) Constraints, Limitations, Assumptions According to the requirements of ADEOS-II/GLI/CTSK1a and ADEOS-II/GLI/CTSK1b, an algorithm has been developed for cloudy/clear discrimination and snow/sea-ice discrimination. The cloudy/clear discriminator consists of a daytime as well as a nighttime mode. The BT 11 BT 3.7 test has been successfully and effectively used in the daytime mode of the cloudy/clear discriminator instead of NDSI technique due to the lack of 1.64 µm channel. The 1.38 µm channel test, which is important for the detection of thin cirrus, has been included in the daytime mode. In the nighttime mode, the BT 8.6 BT 12 brightness temperature difference test based on the different refractive indexes for ice and water is available to detect cirrus cloud. During polar winter, a temperature inversion test is included into both the daytime and the nighttime modes to improve the capability of the cloud detection. The snow/sea-ice discriminator is a daytime only mode. The so-called NDII technique and the reflectance-ratio approach have been used in the algorithm. GLI channel 24 (1.05 µm) is an important channel for discriminating snow from bare sea ice. The rapid decline of ice reflectance in the µm region is the basis for the snow/sea-ice discriminator algorithm. Applications to the MAS data set reveal that the NDII is an effective method for the separation of sea-ice and snow/open-water. The basic features of the algorithm may be summarized as follows: (i) The cloud detection algorithm is designed to work independently without considering surface conditions and to be valid for both polar regions and snow-covered mid-latitude regions. The threshold for each test applied in this algorithm has the same value for GLI Algorithm Description

27 different surfaces. Thus, the cloud-contaminated pixels can be detected before the surface condition is known. (ii) The algorithm is based on pre-defined thresholds. The results of the cloudy/clear discriminator are indicated as 4 levels of confidence with regard to whether a pixel is considered to be cloudy. This algorithm is constructed to be cloud-conservative, minimizing false cloud detection but missing clouds that resemble clear conditions. (iii) The snow/sea-ice discriminator works only during daytime. Fortunately, bare sea-ice occurs primarily in summer. Bright polar summers will offer the opportunity to use the snow/sea-ice discriminator for extended periods of time. (iv) The nighttime mode is necessary during polar winter. The thresholds of the infrared channels used in the nighttime mode should be determined based on nighttime observational data. Due to the lack of nighttime data, the nighttime mode will be validated in the future. (v) Because GLI channels are not exactly matched by MAS channels, the thresholds for each test should be further adjusted to fit the GLI channels. This work has been completed based on existing approaches. However, there are several remaining issues that need to be investigated before and after the launch of ADEOS-II. Some tests in this algorithm should be further validated with upcoming satellite observations and/or higher resolution aircraft data. Our validation relies heavily on the use of observational data, primarily the MAS data. MODIS data will also be considered for validation when such data become available. Since the spectral distribution of the GLI and the MODIS channels are quite similar, we expect MODIS data to be very useful for testing existing approaches to the cloudy/clear discriminator and snow/sea-ice discriminator and for developing new ones as deemed necessary. (6) Publications and Papers Li, W., X. Xiong, K. Stamnes, and B. Chen, An algorithm for mapping snow/ice cover in the Arctic ocean using MODIS Airborne Simulator (MAS) data, 49th Arctic Science Conference, Alaska, October 25-28, (in press). GLI Algorithm Description

### DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team

DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team Steve Ackerman 1, Kathleen Strabala 1, Paul Menzel 1,2, Richard Frey 1, Chris Moeller 1,

### Discriminating clear sky from clouds with MODIS

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 103, NO. D24, PAGES 32,141-32,157, DECEMBER 27, 1998 Discriminating clear sky from clouds with MODIS Steven A. Ackerman, Kathleen I. Strabala, 2 W. Paul Menzel, 3

### DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team

DISCRIMINATING CLEAR-SKY FROM CLOUD WITH MODIS ALGORITHM THEORETICAL BASIS DOCUMENT (MOD35) MODIS Cloud Mask Team Steve Ackerman, Richard Frey, Kathleen Strabala, Yinghui Liu, Liam Gumley, Bryan Baum,

### Comparison of NOAA's Operational AVHRR Derived Cloud Amount to other Satellite Derived Cloud Climatologies.

Comparison of NOAA's Operational AVHRR Derived Cloud Amount to other Satellite Derived Cloud Climatologies. Sarah M. Thomas University of Wisconsin, Cooperative Institute for Meteorological Satellite Studies

### clear sky from clouds with MODIS

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 103, NO. D24, PAGES 32,141-32,157, DECEMBER 27, 1998 Discriminating clear sky from clouds with MODIS Steven A. Ackerman, Kathleen I. Strabala,* W. Paul Menzel,3 Richard

### Cloud Masking and Cloud Products

Cloud Masking and Cloud Products MODIS Operational Algorithm MOD35 Paul Menzel, Steve Ackerman, Richard Frey, Kathy Strabala, Chris Moeller, Liam Gumley, Bryan Baum MODIS Cloud Masking Often done with

### Studying cloud properties from space using sounder data: A preparatory study for INSAT-3D

Studying cloud properties from space using sounder data: A preparatory study for INSAT-3D Munn V. Shukla and P. K. Thapliyal Atmospheric Sciences Division Atmospheric and Oceanic Sciences Group Space Applications

### Cloud detection and clearing for the MOPITT instrument

Cloud detection and clearing for the MOPITT instrument Juying Warner, John Gille, David P. Edwards and Paul Bailey National Center for Atmospheric Research, Boulder, Colorado ABSTRACT The Measurement Of

### Evaluation of the Effect of Upper-Level Cirrus Clouds on Satellite Retrievals of Low-Level Cloud Droplet Effective Radius

Evaluation of the Effect of Upper-Level Cirrus Clouds on Satellite Retrievals of Low-Level Cloud Droplet Effective Radius F.-L. Chang and Z. Li Earth System Science Interdisciplinary Center University

### Clear Sky Radiance (CSR) Product from MTSAT-1R. UESAWA Daisaku* Abstract

Clear Sky Radiance (CSR) Product from MTSAT-1R UESAWA Daisaku* Abstract The Meteorological Satellite Center (MSC) has developed a Clear Sky Radiance (CSR) product from MTSAT-1R and has been disseminating

### Authors: Thierry Phulpin, CNES Lydie Lavanant, Meteo France Claude Camy-Peyret, LPMAA/CNRS. Date: 15 June 2005

Comments on the number of cloud free observations per day and location- LEO constellation vs. GEO - Annex in the final Technical Note on geostationary mission concepts Authors: Thierry Phulpin, CNES Lydie

### Use of ARM/NSA Data to Validate and Improve the Remote Sensing Retrieval of Cloud and Surface Properties in the Arctic from AVHRR Data

Use of ARM/NSA Data to Validate and Improve the Remote Sensing Retrieval of Cloud and Surface Properties in the Arctic from AVHRR Data X. Xiong QSS Group, Inc. National Oceanic and Atmospheric Administration

### Measurement of the effect of biomass burning aerosol on inhibition of cloud formation over the Amazon

Supporting Online Material for Koren et al. Measurement of the effect of biomass burning aerosol on inhibition of cloud formation over the Amazon 1. MODIS new cloud detection algorithm The operational

### Validating MOPITT Cloud Detection Techniques with MAS Images

Validating MOPITT Cloud Detection Techniques with MAS Images Daniel Ziskin, Juying Warner, Paul Bailey, John Gille National Center for Atmospheric Research, P.O. Box 3000, Boulder, CO 80307 ABSTRACT The

### AATSR Technical Note. Improvements to the AATSR IPF relating to Land Surface Temperature Retrieval and Cloud Clearing over Land

AATSR Technical Note Improvements to the AATSR IPF relating to Land Surface Temperature Retrieval and Cloud Clearing over Land Author: Andrew R. Birks RUTHERFORD APPLETON LABORATORY Chilton, Didcot, Oxfordshire

### Towards assimilating IASI satellite observations over cold surfaces - the cloud detection aspect

Towards assimilating IASI satellite observations over cold surfaces - the cloud detection aspect Tuuli Perttula, FMI + Thanks to: Nadia Fourrié, Lydie Lavanant, Florence Rabier and Vincent Guidard, Météo

### Clouds and the Energy Cycle

August 1999 NF-207 The Earth Science Enterprise Series These articles discuss Earth's many dynamic processes and their interactions Clouds and the Energy Cycle he study of clouds, where they occur, and

### Volcanic Ash Monitoring: Product Guide

Doc.No. Issue : : EUM/TSS/MAN/15/802120 v1a EUMETSAT Eumetsat-Allee 1, D-64295 Darmstadt, Germany Tel: +49 6151 807-7 Fax: +49 6151 807 555 Date : 2 June 2015 http://www.eumetsat.int WBS/DBS : EUMETSAT

### The Surface Energy Budget

The Surface Energy Budget The radiation (R) budget Shortwave (solar) Radiation Longwave Radiation R SW R SW α α = surface albedo R LW εσt 4 ε = emissivity σ = Stefan-Boltzman constant T = temperature Subsurface

### Overview of the IR channels and their applications

Ján Kaňák Slovak Hydrometeorological Institute Jan.kanak@shmu.sk Overview of the IR channels and their applications EUMeTrain, 14 June 2011 Ján Kaňák, SHMÚ 1 Basics in satellite Infrared image interpretation

### ECMWF Aerosol and Cloud Detection Software. User Guide. version 1.2 20/01/2015. Reima Eresmaa ECMWF

ECMWF Aerosol and Cloud User Guide version 1.2 20/01/2015 Reima Eresmaa ECMWF This documentation was developed within the context of the EUMETSAT Satellite Application Facility on Numerical Weather Prediction

### Received in revised form 24 March 2004; accepted 30 March 2004

Remote Sensing of Environment 91 (2004) 237 242 www.elsevier.com/locate/rse Cloud detection in Landsat imagery of ice sheets using shadow matching technique and automatic normalized difference snow index

### Assessing Cloud Spatial and Vertical Distribution with Infrared Cloud Analyzer

Assessing Cloud Spatial and Vertical Distribution with Infrared Cloud Analyzer I. Genkova and C. N. Long Pacific Northwest National Laboratory Richland, Washington T. Besnard ATMOS SARL Le Mans, France

### Data Processing Flow Chart

Legend Start V1 V2 V3 Completed Version 2 Completion date Data Processing Flow Chart Data: Download a) AVHRR: 1981-1999 b) MODIS:2000-2010 c) SPOT : 1998-2002 No Progressing Started Did not start 03/12/12

### SAFNWC/MSG Cloud type/height. Application for fog/low cloud situations

SAFNWC/MSG Cloud type/height. Application for fog/low cloud situations 22 September 2011 Hervé LE GLEAU, Marcel DERRIEN Centre de météorologie Spatiale. Lannion Météo-France 1 Fog or low level clouds?

### How to Use the NOAA Enterprise Cloud Mask (ECM) Andrew Heidinger, Tom Kopp, Denis Botambekov and William Straka JPSS Cloud Team August 29, 2015

How to Use the NOAA Enterprise Cloud Mask (ECM) Andrew Heidinger, Tom Kopp, Denis Botambekov and William Straka JPSS Cloud Team August 29, 2015 Outline Describe ECM and its differences to VCM Describe

### Data processing (3) Cloud and Aerosol Imager (CAI)

Data processing (3) Cloud and Aerosol Imager (CAI) 1) Nobuyuki Kikuchi, 2) Haruma Ishida, 2) Takashi Nakajima, 3) Satoru Fukuda, 3) Nick Schutgens, 3) Teruyuki Nakajima 1) National Institute for Environmental

### CERES Edition 2 & Edition 3 Cloud Cover, Cloud Altitude and Temperature

CERES Edition 2 & Edition 3 Cloud Cover, Cloud Altitude and Temperature S. Sun-Mack 1, P. Minnis 2, Y. Chen 1, R. Smith 1, Q. Z. Trepte 1, F. -L. Chang, D. Winker 2 (1) SSAI, Hampton, VA (2) NASA Langley

### Generation of Cloud-free Imagery Using Landsat-8

Generation of Cloud-free Imagery Using Landsat-8 Byeonghee Kim 1, Youkyung Han 2, Yonghyun Kim 3, Yongil Kim 4 Department of Civil and Environmental Engineering, Seoul National University (SNU), Seoul,

### SYNERGISTIC USE OF IMAGER WINDOW OBSERVATIONS FOR CLOUD- CLEARING OF SOUNDER OBSERVATION FOR INSAT-3D

SYNERGISTIC USE OF IMAGER WINDOW OBSERVATIONS FOR CLOUD- CLEARING OF SOUNDER OBSERVATION FOR INSAT-3D ABSTRACT: Jyotirmayee Satapathy*, P.K. Thapliyal, M.V. Shukla, C. M. Kishtawal Atmospheric and Oceanic

### MOD09 (Surface Reflectance) User s Guide

MOD09 (Surface ) User s Guide MODIS Land Surface Science Computing Facility Principal Investigator: Dr. Eric F. Vermote Web site: http://modis-sr.ltdri.org Correspondence e-mail address: mod09@ltdri.org

### RESULTS FROM A SIMPLE INFRARED CLOUD DETECTOR

RESULTS FROM A SIMPLE INFRARED CLOUD DETECTOR A. Maghrabi 1 and R. Clay 2 1 Institute of Astronomical and Geophysical Research, King Abdulaziz City For Science and Technology, P.O. Box 6086 Riyadh 11442,

### The APOLLO cloud product statistics Web service

The APOLLO cloud product statistics Web service Introduction DLR and Transvalor are preparing a new Web service to disseminate the statistics of the APOLLO cloud physical parameters as a further help in

### Labs in Bologna & Potenza Menzel. Lab 3 Interrogating AIRS Data and Exploring Spectral Properties of Clouds and Moisture

Labs in Bologna & Potenza Menzel Lab 3 Interrogating AIRS Data and Exploring Spectral Properties of Clouds and Moisture Figure 1: High resolution atmospheric absorption spectrum and comparative blackbody

### CLOUD CLASSIFICATION EXTRACTED FROM AVHRR AND GOES IMAGERY. M.Derrien, H.Le Gléau

CLOUD CLASSIFICATION EXTRACTED FROM AVHRR AND GOES IMAGERY M.Derrien, H.Le Gléau Météo-France / SCEM / Centre de Météorologie Spatiale BP 147 22302 Lannion. France ABSTRACT We developed an automated pixel-scale

### Spectral Response for DigitalGlobe Earth Imaging Instruments

Spectral Response for DigitalGlobe Earth Imaging Instruments IKONOS The IKONOS satellite carries a high resolution panchromatic band covering most of the silicon response and four lower resolution spectral

### ESCI 107/109 The Atmosphere Lesson 2 Solar and Terrestrial Radiation

ESCI 107/109 The Atmosphere Lesson 2 Solar and Terrestrial Radiation Reading: Meteorology Today, Chapters 2 and 3 EARTH-SUN GEOMETRY The Earth has an elliptical orbit around the sun The average Earth-Sun

### Passive Remote Sensing of Clouds from Airborne Platforms

Passive Remote Sensing of Clouds from Airborne Platforms Why airborne measurements? My instrument: the Solar Spectral Flux Radiometer (SSFR) Some spectrometry/radiometry basics How can we infer cloud properties

### ARM SWS to study cloud drop size within the clear-cloud transition zone

ARM SWS to study cloud drop size within the clear-cloud transition zone (GSFC) Yuri Knyazikhin Boston University Christine Chiu University of Reading Warren Wiscombe GSFC Thanks to Peter Pilewskie (UC)

### WATER BODY EXTRACTION FROM MULTI SPECTRAL IMAGE BY SPECTRAL PATTERN ANALYSIS

WATER BODY EXTRACTION FROM MULTI SPECTRAL IMAGE BY SPECTRAL PATTERN ANALYSIS Nguyen Dinh Duong Department of Environmental Information Study and Analysis, Institute of Geography, 18 Hoang Quoc Viet Rd.,

### GOES-R AWG Cloud Team: ABI Cloud Height

GOES-R AWG Cloud Team: ABI Cloud Height June 8, 2010 Presented By: Andrew Heidinger 1 1 NOAA/NESDIS/STAR 1 Outline Executive Summary Algorithm Description ADEB and IV&V Response Summary Requirements Specification

### Comparison between current and future environmental satellite imagers on cloud classification using MODIS

Remote Sensing of Environment 108 (2007) 311 326 www.elsevier.com/locate/rse Comparison between current and future environmental satellite imagers on cloud classification using MODIS Zhenglong Li a,, Jun

### MODIS IMAGES RESTORATION FOR VNIR BANDS ON FIRE SMOKE AFFECTED AREA

MODIS IMAGES RESTORATION FOR VNIR BANDS ON FIRE SMOKE AFFECTED AREA Li-Yu Chang and Chi-Farn Chen Center for Space and Remote Sensing Research, National Central University, No. 300, Zhongda Rd., Zhongli

### A remote sensing instrument collects information about an object or phenomenon within the

Satellite Remote Sensing GE 4150- Natural Hazards Some slides taken from Ann Maclean: Introduction to Digital Image Processing Remote Sensing the art, science, and technology of obtaining reliable information

### Outline of RGB Composite Imagery

Outline of RGB Composite Imagery Data Processing Division, Data Processing Department Meteorological Satellite Center (MSC) JMA Akihiro SHIMIZU 29 September, 2014 Updated 6 July, 2015 1 Contents What s

### How Landsat Images are Made

How Landsat Images are Made Presentation by: NASA s Landsat Education and Public Outreach team June 2006 1 More than just a pretty picture Landsat makes pretty weird looking maps, and it isn t always easy

### VIIRS-CrIS mapping. NWP SAF AAPP VIIRS-CrIS Mapping

NWP SAF AAPP VIIRS-CrIS Mapping This documentation was developed within the context of the EUMETSAT Satellite Application Facility on Numerical Weather Prediction (NWP SAF), under the Cooperation Agreement

### ATM S 111, Global Warming: Understanding the Forecast

ATM S 111, Global Warming: Understanding the Forecast DARGAN M. W. FRIERSON DEPARTMENT OF ATMOSPHERIC SCIENCES DAY 1: OCTOBER 1, 2015 Outline How exactly the Sun heats the Earth How strong? Important concept

### The APOLLO cloud product statistics Web service The APOLLO cloud product statistics Web service

The APOLLO cloud product statistics Web service Introduction DLR and Transvalor are preparing a new Web service to disseminate the statistics of the APOLLO cloud physical parameters as a further help in

### CLOUD MASKING AND CLOUD PRODUCTS ROUNDTABLE EXPECTED PARTICIPANTS: ACKERMAN, HALL, WAN, VERMOTE, BARKER, HUETE, BROWN, GORDON, KAUFMAN, SCHAAF, BAUM

CLOUD MASKING AND CLOUD PRODUCTS ROUNDTABLE EXPECTED PARTICIPANTS: ACKERMAN, HALL, WAN, VERMOTE, BARKER, HUETE, BROWN, GORDON, KAUFMAN, SCHAAF, BAUM NOMINAL PURPOSE: DISCUSSION OF TESTS FOR ACCURACY AND

### Global Moderate Resolution Imaging Spectroradiometer (MODIS) cloud detection and height evaluation using CALIOP

Click Here for Full Article JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 113,, doi:10.1029/2008jd009837, 2008 Global Moderate Resolution Imaging Spectroradiometer (MODIS) cloud detection and height evaluation

### GLOBAL FORUM London, October 24 & 25, 2012

GLOBAL FORUM London, October 24 & 25, 2012-1 - Global Observations of Gas Flares Improving Global Observations of Gas Flares With Data From the Suomi NPP Visible Infrared Imaging Radiometer Suite (VIIRS)

### Remote sensing of cloud cover in the Arctic region from AVHRR data during the ARTIST experiment

INT. J. REMOTE SENSING, 2003, VOL. 24, NO. 3, 437 456 Remote sensing of cloud cover in the Arctic region from AVHRR data during the ARTIST experiment C. ANANASSO, R SANTOLERI, S. MARULLO and F. D ORTENZIO

### Evaluation of VIIRS cloud top property climate data records and their potential improvement with CrIS

Evaluation of VIIRS cloud top property climate data records and their potential improvement with CrIS Dr. Bryan A. Baum (PI) Space Science and Engineering Center University of Wisconsin-Madison Madison,

### Seasonal & Daily Temperatures. Seasons & Sun's Distance. Solstice & Equinox. Seasons & Solar Intensity

Seasonal & Daily Temperatures Seasons & Sun's Distance The role of Earth's tilt, revolution, & rotation in causing spatial, seasonal, & daily temperature variations Please read Chapter 3 in Ahrens Figure

### Satellite remote sensing using AVHRR, ATSR, MODIS, METEOSAT, MSG

Satellite remote sensing using AVHRR, ATSR, MODIS, METEOSAT, MSG Ralf Meerkötter, Luca Bugliaro, Knut Dammann, Gerhard Gesell, Christine König, Waldemar Krebs, Hermann Mannstein, Bernhard Mayer, presented

### Night Microphysics RGB Nephanalysis in night time

Copyright, JMA Night Microphysics RGB Nephanalysis in night time Meteorological Satellite Center, JMA What s Night Microphysics RGB? R : B15(I2 12.3)-B13(IR 10.4) Range : -4 2 [K] Gamma : 1.0 G : B13(IR

### STAR Algorithm and Data Products (ADP) Beta Review. Suomi NPP Surface Reflectance IP ARP Product

STAR Algorithm and Data Products (ADP) Beta Review Suomi NPP Surface Reflectance IP ARP Product Alexei Lyapustin Surface Reflectance Cal Val Team 11/26/2012 STAR ADP Surface Reflectance ARP Team Member

### A SURVEY OF CLOUD COVER OVER MĂGURELE, ROMANIA, USING CEILOMETER AND SATELLITE DATA

Romanian Reports in Physics, Vol. 66, No. 3, P. 812 822, 2014 ATMOSPHERE PHYSICS A SURVEY OF CLOUD COVER OVER MĂGURELE, ROMANIA, USING CEILOMETER AND SATELLITE DATA S. STEFAN, I. UNGUREANU, C. GRIGORAS

### FRESCO. Product Specification Document FRESCO. Authors : P. Wang, R.J. van der A (KNMI) REF : TEM/PSD2/003 ISSUE : 3.0 DATE : 30.05.

PAGE : 1/11 TITLE: Product Specification Authors : P. Wang, R.J. van der A (KNMI) PAGE : 2/11 DOCUMENT STATUS SHEET Issue Date Modified Items / Reason for Change 0.9 19.01.06 First Version 1.0 22.01.06

### Supervised Classification workflow in ENVI 4.8 using WorldView-2 imagery

Supervised Classification workflow in ENVI 4.8 using WorldView-2 imagery WorldView-2 is the first commercial high-resolution satellite to provide eight spectral sensors in the visible to near-infrared

### Radiative effects of clouds, ice sheet and sea ice in the Antarctic

Snow and fee Covers: Interactions with the Atmosphere and Ecosystems (Proceedings of Yokohama Symposia J2 and J5, July 1993). IAHS Publ. no. 223, 1994. 29 Radiative effects of clouds, ice sheet and sea

### Scientific Basis and Initial Evaluation of the CLAVR-1 Global Clear/Cloud Classification Algorithm for the Advanced Very High Resolution Radiometer

656 JOURNAL OF ATMOSPHERIC AND OCEANIC TECHNOLOGY VOLUME 16 Scientific Basis and Initial Evaluation of the CLAVR-1 Global Clear/Cloud Classification Algorithm for the Advanced Very High Resolution Radiometer

### Obtaining and Processing MODIS Data

Obtaining and Processing MODIS Data MODIS is an extensive program using sensors on two satellites that each provide complete daily coverage of the earth. The data have a variety of resolutions; spectral,

### Review for Introduction to Remote Sensing: Science Concepts and Technology

Review for Introduction to Remote Sensing: Science Concepts and Technology Ann Johnson Associate Director ann@baremt.com Funded by National Science Foundation Advanced Technological Education program [DUE

### 2.3 Spatial Resolution, Pixel Size, and Scale

Section 2.3 Spatial Resolution, Pixel Size, and Scale Page 39 2.3 Spatial Resolution, Pixel Size, and Scale For some remote sensing instruments, the distance between the target being imaged and the platform,

### Active Fire Monitoring: Product Guide

Active Fire Monitoring: Product Guide Doc.No. Issue : : EUM/TSS/MAN/15/801989 v1c EUMETSAT Eumetsat-Allee 1, D-64295 Darmstadt, Germany Tel: +49 6151 807-7 Fax: +49 6151 807 555 Date : 14 April 2015 http://www.eumetsat.int

### Cloud Climatology for New Zealand and Implications for Radiation Fields

Cloud Climatology for New Zealand and Implications for Radiation Fields G. Pfister, R.L. McKenzie, J.B. Liley, A. Thomas National Institute of Water and Atmospheric Research, Lauder, New Zealand M.J. Uddstrom

### Hyperspectral Satellite Imaging Planning a Mission

Hyperspectral Satellite Imaging Planning a Mission Victor Gardner University of Maryland 2007 AIAA Region 1 Mid-Atlantic Student Conference National Institute of Aerospace, Langley, VA Outline Objective

### USING THE GOES 3.9 µm SHORTWAVE INFRARED CHANNEL TO TRACK LOW-LEVEL CLOUD-DRIFT WINDS ABSTRACT

USING THE GOES 3.9 µm SHORTWAVE INFRARED CHANNEL TO TRACK LOW-LEVEL CLOUD-DRIFT WINDS Jason P. Dunion 1 and Christopher S. Velden 2 1 NOAA/AOML/Hurricane Research Division, 2 UW/CIMSS ABSTRACT Low-level

### The Next Generation Flux Analysis: Adding Clear-Sky LW and LW Cloud Effects, Cloud Optical Depths, and Improved Sky Cover Estimates

The Next Generation Flux Analysis: Adding Clear-Sky LW and LW Cloud Effects, Cloud Optical Depths, and Improved Sky Cover Estimates C. N. Long Pacific Northwest National Laboratory Richland, Washington

### Advances in Cloud Imager Remote Sensing

Advances in Cloud Imager Remote Sensing Andrew Heidinger NOAA/NESDIS/ORA Madison, Wisconsin With material from Mike Pavolonis, Robert Holz, Amato Evan and Fred Nagle STAR Science Symposium November 9,

### VALIDATION OF SAFNWC / MSG CLOUD PRODUCTS WITH ONE YEAR OF SEVIRI DATA

VALIDATION OF SAFNWC / MSG CLOUD PRODUCTS WITH ONE YEAR OF SEVIRI DATA M.Derrien 1, H.Le Gléau 1, Jean-François Daloze 2, Martial Haeffelin 2 1 Météo-France / DP / Centre de Météorologie Spatiale. BP 50747.

### GOES-R Advanced Baseline Imager (ABI) Algorithm Theoretical Basis Document For Low Cloud and Fog

NOAA NESDIS CENTER for SATELLITE APPLICATIONS and RESEARCH GOES-R Advanced Baseline Imager (ABI) Algorithm Theoretical Basis Document For Low Cloud and Fog Corey Calvert, UW/CIMSS Mike Pavolonis, NOAA/NESDIS/STAR

### Solar Flux and Flux Density. Lecture 3: Global Energy Cycle. Solar Energy Incident On the Earth. Solar Flux Density Reaching Earth

Lecture 3: Global Energy Cycle Solar Flux and Flux Density Planetary energy balance Greenhouse Effect Vertical energy balance Latitudinal energy balance Seasonal and diurnal cycles Solar Luminosity (L)

### Total radiative heating/cooling rates.

Lecture. Total radiative heating/cooling rates. Objectives:. Solar heating rates.. Total radiative heating/cooling rates in a cloudy atmosphere.. Total radiative heating/cooling rates in different aerosol-laden

### Selecting the appropriate band combination for an RGB image using Landsat imagery

Selecting the appropriate band combination for an RGB image using Landsat imagery Ned Horning Version: 1.0 Creation Date: 2004-01-01 Revision Date: 2004-01-01 License: This document is licensed under a

### Cloud Masking for Remotely Sensed Data Using Spectral and Principal Components Analysis

ETASR - Engineering, Technology & Applied Science Research Vol. 2, o. 3, 2012, 221-225 221 Cloud Masking for Remotely Sensed Data Using Spectral and Principal Components Analysis Asmala Ahmad Faculty of

### P1.21 GOES CLOUD DETECTION AT THE GLOBAL HYDROLOGY AND CLIMATE CENTER

P1.21 GOES CLOUD DETECTION AT THE GLOBAL HYDROLOGY AND CLIMATE CENTER Gary J. Jedlovec* NASA/MSFC/Global Hydrology and Climate Center National Space Science and Technology Center Huntsville, Alabama and

### Best practices for RGB compositing of multi-spectral imagery

Best practices for RGB compositing of multi-spectral imagery User Service Division, EUMETSAT Introduction Until recently imagers on geostationary satellites were limited to 2-3 spectral channels, i.e.

### ES 106 Laboratory # 5 EARTH-SUN RELATIONS AND ATMOSPHERIC HEATING

ES 106 Laboratory # 5 EARTH-SUN RELATIONS AND ATMOSPHERIC HEATING 5-1 Introduction Weather is the state of the atmosphere at a particular place for a short period of time. The condition of the atmosphere

### Observed Cloud Cover Trends and Global Climate Change. Joel Norris Scripps Institution of Oceanography

Observed Cloud Cover Trends and Global Climate Change Joel Norris Scripps Institution of Oceanography Increasing Global Temperature from www.giss.nasa.gov Increasing Greenhouse Gases from ess.geology.ufl.edu

### Landsat 7 Automatic Cloud Cover Assessment

Landsat 7 Automatic Cloud Cover Assessment Richard R. Irish Science Systems and Applications, Inc. NASA s Goddard Space Flight Center, Greenbelt, Maryland ABSTRACT An automatic cloud cover assessment algorithm

### Multiangle cloud remote sensing from

Multiangle cloud remote sensing from POLDER3/PARASOL Cloud phase, optical thickness and albedo F. Parol, J. Riedi, S. Zeng, C. Vanbauce, N. Ferlay, F. Thieuleux, L.C. Labonnote and C. Cornet Laboratoire

### Remote Sensing of Clouds from Polarization

Remote Sensing of Clouds from Polarization What polarization can tell us about clouds... and what not? J. Riedi Laboratoire d'optique Atmosphérique University of Science and Technology Lille / CNRS FRANCE

### A comparison of NOAA/AVHRR derived cloud amount with MODIS and surface observation

A comparison of NOAA/AVHRR derived cloud amount with MODIS and surface observation LIU Jian YANG Xiaofeng and CUI Peng National Satellite Meteorological Center, CMA, CHINA outline 1. Introduction 2. Data

### Technical note on MISR Cloud-Top-Height Optical-depth (CTH-OD) joint histogram product

Technical note on MISR Cloud-Top-Height Optical-depth (CTH-OD) joint histogram product 1. Intend of this document and POC 1.a) General purpose The MISR CTH-OD product contains 2D histograms (joint distributions)

### How To Understand Cloud Properties From Satellite Imagery

P1.70 NIGHTTIME RETRIEVAL OF CLOUD MICROPHYSICAL PROPERTIES FOR GOES-R Patrick W. Heck * Cooperative Institute for Meteorological Satellite Studies, University of Wisconsin-Madison Madison, Wisconsin P.

### Evaluations of the CALIPSO Cloud Optical Depth Algorithm Through Comparisons with a GOES Derived Cloud Analysis

Generated using V3.0 of the official AMS LATEX template Evaluations of the CALIPSO Cloud Optical Depth Algorithm Through Comparisons with a GOES Derived Cloud Analysis Katie Carbonari, Heather Kiley, and

### Chapter 2: Solar Radiation and Seasons

Chapter 2: Solar Radiation and Seasons Spectrum of Radiation Intensity and Peak Wavelength of Radiation Solar (shortwave) Radiation Terrestrial (longwave) Radiations How to Change Air Temperature? Add

### Interactive comment on Total cloud cover from satellite observations and climate models by P. Probst et al.

Interactive comment on Total cloud cover from satellite observations and climate models by P. Probst et al. Anonymous Referee #1 (Received and published: 20 October 2010) The paper compares CMIP3 model

### NOAA NESDIS CENTER for SATELLITE APPLICATIONS and RESEARCH ALGORITHM THEORETICAL BASIS DOCUMENT. ABI Cloud Mask

NOAA NESDIS CENTER for SATELLITE APPLICATIONS and RESEARCH ALGORITHM THEORETICAL BASIS DOCUMENT ABI Cloud Mask Andrew Heidinger, NOAA/NESDIS/STAR William C. Straka III, SSEC/CIMSS Version 3.0 June 11,

### Resolutions of Remote Sensing

Resolutions of Remote Sensing 1. Spatial (what area and how detailed) 2. Spectral (what colors bands) 3. Temporal (time of day/season/year) 4. Radiometric (color depth) Spatial Resolution describes how

### Energy Pathways in Earth s Atmosphere

BRSP - 10 Page 1 Solar radiation reaching Earth s atmosphere includes a wide spectrum of wavelengths. In addition to visible light there is radiation of higher energy and shorter wavelength called ultraviolet

### SAMPLE MIDTERM QUESTIONS

Geography 309 Sample MidTerm Questions Page 1 SAMPLE MIDTERM QUESTIONS Textbook Questions Chapter 1 Questions 4, 5, 6, Chapter 2 Questions 4, 7, 10 Chapter 4 Questions 8, 9 Chapter 10 Questions 1, 4, 7

### 163 ANALYSIS OF THE URBAN HEAT ISLAND EFFECT COMPARISON OF GROUND-BASED AND REMOTELY SENSED TEMPERATURE OBSERVATIONS

ANALYSIS OF THE URBAN HEAT ISLAND EFFECT COMPARISON OF GROUND-BASED AND REMOTELY SENSED TEMPERATURE OBSERVATIONS Rita Pongrácz *, Judit Bartholy, Enikő Lelovics, Zsuzsanna Dezső Eötvös Loránd University,

### A review of cloud top height and optical depth histograms from MISR, ISCCP, and MODIS

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 115,, doi:10.1029/2009jd013422, 2010 A review of cloud top height and optical depth histograms from MISR, ISCCP, and MODIS Roger Marchand, 1 Thomas Ackerman, 1 Mike

### REMOTE SENSING OF CLOUD-AEROSOL RADIATIVE EFFECTS FROM SATELLITE DATA: A CASE STUDY OVER THE SOUTH OF PORTUGAL

REMOTE SENSING OF CLOUD-AEROSOL RADIATIVE EFFECTS FROM SATELLITE DATA: A CASE STUDY OVER THE SOUTH OF PORTUGAL D. Santos (1), M. J. Costa (1,2), D. Bortoli (1,3) and A. M. Silva (1,2) (1) Évora Geophysics