Figure 2 - uploaded by Hitesh Arvindbhai Solanki
Content may be subject to copyright.
Density plots of the satellite SST validation against (a) the IMB in-situ SST measurements and (b) the iQuam in-situ SST measurements.
Source publication
The INSAT-3D imager (4 km) and Moderate Resolution Imaging Spectroradiometer (MODIS) sensor on-board Aqua and Terra space-platforms level-2 (1 km) sea surface temperature (SSTskin) product accuracy has been analysed over waters surrounding the Indian subcontinent by indirect comparison method using collocated bulk in-situ measurements (SSTdepth) fo...
Contexts in source publication
Context 1
... general, the retrieval error (RMSE) of all the three satellites w.r.t. in-situ observations is in the range of around 0.60-0.70°C. Figures 2(a,b) show combined density plots for all the 3 years (October 2013-October 2016) indicating overall statistics of satellite valida- tion against in-situ measurements. The retrieval error is observed to be slightly higher in case of satellite SST validation against iQuam data set (RMSE ranging in 0.67-0.70°C) as compared to IMB in-situ SST data set (RMSE ranging in 0.60-0.66°C). The value of coefficient of determination (R 2 ) is also lower in case of validation against iQuam data set (0.55-0.71) than that against IMB data set (0.64-0.76). Like satellite observations, in- situ SST observations are also never fully accurate (Feng and Ignatov 2010) and are subjected to variations in their accuracy. However, the accuracy of buoy SST observa- tions is usually better than 0.5°C (Reynolds 2001) which in turn is much better than the SST measurements taken through ships having typical RMSE larger than 1°C (Kent, Challenor, and Taylor 1999). The use of ship-borne measurements may introduce errors of depth variations and may impact the accuracy of the SST retrievals ( Donlon et al. 2002). In order to confirm the accuracy of iQuam SST w.r.t. IMB SST, the former in-situ data set was collocated within a spatial window of ±1 km and temporal window of ±0.5 h of particular IMB SST observation. Although both in-situ SST data sets (IMB SST and iQuam SST) are found to be in good agreement (R 2 = 0.92) showing RMSE of about 0.34°C with a negligible bias of less than 0.01°C, use of conventional ships and IMOS ships in case of iQuam could probably lead to slight inaccuracy in the iQuam in-situ SST data set. The additional errors may be induced during transmission and processing stages of in-situ measurements. It has already been mentioned previously that the iQuam data set comprises in-situ observations from a large number of platforms including drifters, moorings, and ships. It is also well known that sensor accuracy differs greatly across these platform types. Therefore, it is imperative to determine the statistics of satellite SST validation against iQuam SST data set associated with each individual platform. Moreover, this exercise would be beneficial to confirm the exact reason behind inaccuracy in the iQuam in-situ SST data set. Table 2 shows combined statistical analysis of satellite SST validation against iQuam SST data set associated with each individual platform for all 3 years. In general, quality-controlled iQuam data set is globally obtained from eight different types of platforms, namely (1) ships (conventional), (2) drifting buoys (conventional drifters), (3) tropical mooring buoys (TMB), (4) coastal mooring buoys (CMB), (5) argo floats, (6) high resolution drifters, (7) IMOS ships, and (8) CRW buoys. Out of these eight platforms, the entire iQuam data set over this study region is found to be associated with only three platforms, namely conventional ships, TMB, and CMB. The statistical results clearly reveal that all three satellite sensors are showing higher RMSE (0.78-0.94°C) and least R 2 (0.32-0.66°C) when validated against iQuam SST data set obtained from conventional ships. On the other hand, satellite SST validation against TMB and CMB data sets is showing comparable results with RMSE ranging from 0.52°C to 0.65°C and R 2 ranging from 0.62°C to 0.79°C. The satellites are found to be underestimating SST obtained from all three platforms. Unlike the INSAT-3D which is having near similar RMSE (0.63-0.78°C) and bias (−0.16°C to −0.20°C) in case of all three in-situ platforms, the MODIS is showing considerably high retrieval error (RMSE = 0.84-0.94°C, bias = −0.20°C to −0.25°C) in case of validation against ship data set as compared to that against the other two platforms (RMSE = 0.52-0.61°C, bias = −0.03°C to −0.10°C). Figure 3 is showing percentage of match-up observations obtained on validation of satellite SST against iQuam SST asso- ciated with individual platform. Since the iQuam in-situ SST data sets obtained from all three platforms are significantly contributing to the validation statistics, therefore col- lective iQuam data set from all three platforms has been used for further study and analysis. Figures 4(a,b) show combined frequency histogram for all 3 years for the temperature difference between satellite observations and in-situ measurements. The temperature difference is in the range of −2.5°C to +2.5°C. In case of almost all three satellites, maximum number of satellite observations is found to be underestimating the in-situ SST observations. Overall, approximately 46-61% of the total collocated satellite SST data set is showing underestimation w.r.t. in-situ SST observation. In particular, around 58-61% observations of INSAT-3D are showing underestimation while in case of MODIS sensor, around 46-58% observations are indicating underestimation. A significant num- ber of INSAT-3D observations are showing temperature difference w.r.t. in-situ observa- tions in the range of −1-0.5°C. On the other hand, maximum number of MODIS observations are showing temperature difference in the range of ...
Context 2
... observations is in the range of around 0.60-0.70°C. Figures 2(a,b) show combined density plots for all the 3 years (October 2013-October 2016) indicating overall statistics of satellite valida- tion against in-situ measurements. The retrieval error is observed to be slightly higher in case of satellite SST validation against iQuam data set (RMSE ranging in 0.67-0.70°C) ...
Similar publications
The spectral estimates of Aerosol Optical Depth (AOD, τ) were made by operating a Microtops-II sun photometer in the spectral range 0.380–0.870 μm over Tripura in north-eastern India and analyzed to infer the aerosol types and source characteristics in different seasons. The Ångström exponent (α) derived from spectral AOD in different wavelength (λ...
Citations
... The Indian National Satellite-3D (INSAT-3D) imager is an improved version of the Very High-Resolution Radiometer (VHRR-2) onboard INSAT-3D, a new generation of stationary satellites in India. The bias between INSAT-3D imager SST and in situ, SST in the Indian Ocean was between −0.16 • C and −0.20 • C, and the bias between the two in the Arabian Ocean was −0.27 • C [30]. The weather imager (MI) is a payload onboard the Communication, Ocean, and Meteorological Satellite (COMS), Korea's first geostationary satellite and can retrieve SST. ...
The Fengyun-4A (FY-4A) satellite is a new-generation geostationary meteorological satellite developed by China. The advanced geosynchronous radiation imager (AGRI), one of the key payloads onboard FY-4A, can monitor sea surface temperature (SST). This paper compares FY-4A/AGRI SST with in situ and Himawari-8/advanced Himawari imager (AHI) SST. The study area spans 30°E–180°E, 60°S–60°N, and the study period is from January 2019 to December 2021. The matching time window of the three data is 30 min, and the space window is 0.1°. The quality control criterion is to select all clear sky and well-distributed matchups within the study period, removing the influence of SST fronts. The results of the difference between FY-4A/AGRI and in situ SST show a bias of −0.12 °C, median of −0.05 °C, standard deviation (STD) of 0.76 °C, robust standard deviation (RSD) of 0.68 °C, and root mean square error (RMSE) of 0.77 °C for daytime and a bias of 0.00 °C, median of 0.05 °C, STD of 0.78 °C, RSD of 0.72 °C, and RMSE of 0.78 °C for nighttime. The results of the difference between FY-4A/AGRI SST and Himawari-8/AHI SST show a bias of 0.04 °C, median of 0.10 °C, STD of 0.78 °C, RSD of 0.70 °C, and RMSE of 0.78 °C for daytime and the bias of 0.30 °C, median of 0.34 °C, STD of 0.81 °C, RSD of 0.76 °C, and RMSE of 0.86 °C for nighttime. The three-way error analysis also indicates a relatively larger error of AGRI SST. Regarding timescale, the bias and STD of FY-4A/AGRI SST show no seasonal correlation, but FY-4A/AGRI SST has a noticeable bias jump in the study period. Regarding spatial scale, FY-4A/AGRI SST shows negative bias at the edge of the AGRI SST coverage in the Pacific region near 160°E longitude and positive bias in high latitudes of the southern hemisphere. The accuracy of FY-4A/AGRI SST depends on the satellite zenith angle and water vapor. Further research on the FY-4A/AGRI SST retrieval algorithm accounting for the variability of water vapor will be conducted.
... In the previous study, the RMSE value for the MODIS sensor was reported as 2.06 (Narayanan et al. 2013) in the BoB using CTD data. Tyagi et al. (2018) ( ...
... Narayanan et al. 2013 found a good correlation (r = 0.93) between MODIS sensors and in-situ data with R 2 = 0.86. A similar result was also found by Tyagi et al. (2018). ...
... If more collocated pair data were obtained, the results would be an effective one. For example, Tyagi et al. (2018) considered the number of observations of 2895 and resulted in the R2 value of 0.76 using the MODIS sensor. ...
This study examined daytime and nighttime sea surface temperature (SST) and its diurnal variability (D-SST) in three regions of the northern Bay of Bengal (BoB) during 2000–2019 by using MODIS Terra satellite data. An increasing trend of SST was observed in the Bangladesh coast, outer coast, and offshore. The trend was ranged around 0.10–0.16 °C per decade during daytime and 0.18–0.27 °C per decade during nighttime. The increased SST in the study area may be attributed due to rapid industrialization and global warming. Monthly SST represented four successive phases, such as warming-cooling and warming-cooling, which started in February. Month wise SST concentrations varied from 29.74 °C to 23.32 °C during nighttime and 29.613 °C to 23.446 °C during daytime in the northern BoB. The comparative SST among three different stations showed that annual offshore SST was higher during daytime and lower during nighttime than the outer & Bangladesh coast. Annual daytime maximum and minimum SST were observed at 28.45 °C and 27.27 °C, respectively. During the nighttime, it was ranged from 28.33 °C to 27.05 °C. The presence of nearby landmass may be the factor behind the variation of SST during day and nighttime. About 65–80% of monthly D-SST and 70–80% of annual D-SST value was recorded below 0.3 °C in our study area. A significantly (p < 0.001) moderate correlation (r = 0.54) was found between the MODIS Terra satellite data and NODC in-situ data.
... INSAT-3D has shown a diurnal variation of 3 • C in the Arabian Sea (AS) buoy location (Mathur et al., 2006). Daily INSAT-3D SST has shown more underestimation in comparison with MODIS observation with a bias of − 0.18 • C (Tyagi et al., 2018). Subsequently, one-Dimensional Variational physical retrieval and Nonlinear SST (NLSST) based algorithms have been applied to INSAT-3D SST retrieval, to account for the sun intrusion effects, and the algorithm has been validated with the SST Quality Monitor (iQuam) observations (Gangwar and Thapliyal, 2020). ...
... On the other hand, under low wind conditions and intense solar radiation, thermal stratification occurs in the upper ocean layer during daytime leading to large deviations between satellite and in-situ measurements (Donlon et al., (2002). This technique has been implemented by many researchers across the globe to validate satellite-derived SST's with in-situ observations (Kim, E.J. et al. 2010, Tyagi et al., 2018, Zhang et al., 2020, Saleh and Al-Anzi, 2021. ...
... In addition, large variations in AODs are observed during monsoon due to the large interannual variations in rainfall over the NIO (Li and Ramanathan, 2002). All these LLCs and AOD variability have a significant influence on the absorption and scattering mechanism of the IR radiation reaching the sensor and hence contribute to the high errors in the SST retrieval (Tyagi et al., 2018). ...
Sea Surface Temperature is an Essential Climate Variable used in oceanographic and meteorological application studies such as air-sea interaction, ocean mixing, boundary layer processes, and ocean state forecast, which require high temporal resolution SST. Indian geostationary satellite INSAT-3D imager has been designed to provide such information at 30 minutes intervals, 4 km spatial resolution along with several other parameters over the North Indian Ocean. In this study, the 30-minute interval INSAT-3D SST and its diurnal variability are validated (inter-compared) with collocated in-situ buoy measurements and compared with contemporary MODerate-resolution Imaging Spectroradiometer SST observations. A reasonable agreement exists between the INSAT-3D and in situ data with a correlation coefficient of 0.70 and Root Mean Square Error (RMSE) of 1.28°C. This relationship is better for the daytime observations and during the pre and post-monsoon seasons, while the relationship is relatively weaker for the nighttime and monsoon seasons with remarked imprints from convective processes in the upper oceans. For the Arabian Sea and Bay of Bengal regions, the INSAT-3D SST shows better correlation and negative bias in comparison with in situ observations, while it is weaker with positive bias for the Equatorial Indian Ocean. Inter-comparison with MODIS Aqua/Terra day and night passes SST also shows a similar relationship, however with a much-improved correlation coefficient (∼0.95) and highly consistent in spatial and temporal variability with a cool bias of 0.14°C. INSAT-3D SST shows a significantly high average diurnal temperature difference of 2.24°C, whereas buoys show 0.54°C. These high RMSE and diurnal variability results suggest that INSAT-3D SST retrieval can be improved to study the many oceanographic processes such as eddies, fronts, and estimation of air-sea fluxes. Therefore, this study helps to improve the retrieval algorithm of INSAT-3D SST, by developing the region-specific regression coefficients along with data merging and assimilation techniques.
... Brown et al. (1999) described in detail the SST retrieval algorithm from MODIS observations. Due to the appropriate spatial and temporal resolution of MODIS images, many studies in different regions have evaluated the SST products of this sensor (Donlon et al. 2002;Minnett et al. 2002;Chan and Gao 2005;Barton 2007;Donlon et al. 2008;Minnett et al. 2011;Castro et al. 2012;Gentemann 2014;Tyagi et al. 2018;Bernardello et al. 2016;Gong and Wong 2018;Tavares et al. 2019;Androulakis et al. 2020). ...
... The depth of water surface temperature measurements by buoys, called bulk-SST, varies from approximately 0.5-5 m (Tyagi et al. 2018). Temporal resolution of buoys SST data is every 10 min. ...
Satellite-based sea surface temperature data is one of the important sources of information for climate and ecological studies, weather forecasting and simulation of atmospheric models. Moderate resolution imaging spectroradiometer sensor onboard the Terra and Aqua satellites provides sea surface temperature products with various temporal and spatial resolutions. In this study, daily 4 km Terra level 3 sea surface temperature products were evaluated using local observations. For this purpose, measurements of National Data Buoy Center standard buoys were used at 14 different points from 2015 to 2018. After extracting high-quality satellite data corresponding to the place and time of buoy measurement, statistical comparison was performed between these two datasets on both day and night. Time series derived from the Terra satellite during the day and night showed an average correlation of 0.93 and 0.96 with buoy measurements. The results showed that Terra level 3 composite for day and night on average determines the Sea Surface Temperature quantity 0.16 C and 0.24 C less than the buoys, respectively. The mean root mean square error of buoy satellite differences during the day and night were 0.76 C and 0.54 C, respectively. Further, in situ measurements and corresponding values obtained from the satellite products were compared in different seasons. In all seasons, the negative bias values of satellite products were higher at night than during the day. On average, the highest error values of Terra sea surface temperature products are related to the summer season, when this parameter reaches its maximum during the year.
... The estimated error over both regions is found to be within the global Chl-a retrieval accuracy of ±35%. Geetika et al. (2018) and Mathur et al. (2006). The statistical tools used to explain the relationship between daily estimated AOD values and Chl-a concentrations include Pearson's correlation coefficient (also known as Pearson r-test) represented by "r" and coefficient of determination (COD), i.e., square of "r", represented by "R 2 ." ...
The paper investigates the impact of atmospheric dust deposition on ocean biological productivity in association with oceanic supply of nutrients over specific regions of the Arabian Sea (20°N, 69°E) and the Bay of Bengal (20°N, 87°E) during wintertime (November–March) from the year 2012 to 2017 using satellite-based observations. During winter, selected regions are characterized by higher Chlorophyll-a (Chl-a) and major oceanic vertical supply of nutrients. Moderate Resolution Imaging Spectroradiometer onboard Aqua space-platform is used to obtain Chl-a and aerosol optical depth (AOD) data. Blended Chl-a daily product from various satellite sensors is also used. There are a total of nine cases (seven cases of the Arabian Sea region and two cases of the Bay of Bengal region) where episodic Chl-a enhancements following high AOD values are observed. Chl-a maxima lag behind AOD maxima by 1 to 4 days. Modern-Era Retrospective analysis for Research and Applications (Version-2) is used for AOD and dust deposition flux estimation. Estimated dust deposition flux ranges between 0.44 and 27.68 mg m⁻² day⁻¹.
... INSAT-3DR (Repeat) with the same payload was launched to 74°E in 2016, with INSAT-3DS (Second Repeat) scheduled for launch in 2022. Comparisons with the iQuam drifter data in the seas around India, revealed mean differences ranging from −0.16 K to −0.20 K, being worse in the Arabian Sea at −0.27 K with greater values during the monsoon season (Tyagi et al., 2018). ...
Sea-surface temperature (SST) was one of the first ocean variables to be studied from earth observation satellites. Pioneering images from infrared scanning radiometers revealed the complexity of the surface temperature fields, but these were derived from radiance measurements at orbital heights and included the effects of the intervening atmosphere. Corrections for the effects of the atmosphere to make quantitative estimates of the SST became possible when radiometers with multiple infrared channels were deployed in 1979. At the same time, imaging microwave radiometers with SST capabilities were also flown. Since then, SST has been derived from infrared and microwave radiometers on polar orbiting satellites and from infrared radiometers on geostationary spacecraft. As the performances of satellite radiometers and SST retrieval algorithms improved, accurate, global, high resolution, frequently sampled SST fields became fundamental to many research and operational activities. Here we provide an overview of the physics of the derivation of SST and the history of the development of satellite instruments over half a century. As demonstrated accuracies increased, they stimulated scientific research into the oceans, the coupled ocean-atmosphere system and the climate. We provide brief overviews of the development of some applications, including the feasibility of generating Climate Data Records. We summarize the important role of the Group for High Resolution SST (GHRSST) in providing a forum for scientists and operational practitioners to discuss problems and results, and to help coordinate activities world-wide, including alignment of data formatting and protocols and research. The challenges of burgeoning data volumes, data distribution and analysis have benefited from simultaneous progress in computing power, high capacity storage, and communications over the Internet, so we summarize the development and current capabilities of data archives. We conclude with an outlook of developments anticipated in the next decade or so.
... According to a validation exercise done by Geetika et al. (2018), MODIS-retrieved SST was observed to have bias ranging between approximately 0.06°C to −0.12°C over the waters surrounding the Indian sub-continent. Fig. 9 shows month-wise analysis of daily SST and upwelling observations. ...
This paper investigates the impact of wind-induced coastal upwelling on bio-optical properties over waters off the Kanyakumari coast for two consecutive winters (October 2011 to March 2012 and October 2012 to March 2013). Level-2 wind datasets from Oceansat-2 Scatterometer (12.5 km) have been used to estimate rate of upwelling using a published approach. Moderate-resolution Imaging Spectroradiometer on-board Aqua and Terra space-platform has been used to obtain bio-optical datasets (level-3, 4 km) that include chlorophyll-a (Chl-a) and particulate organic carbon (POC) concentration, absorption due to phytoplankton at 443 nm (aph_443), absorption due to detritus and gelbstoff at 443 nm (adg_443) and remote sensing reflectance (Rrs) at multiple wavelengths. Wind speed resulting in upwelling is found to be in range of 6.95 m s−1–18.67 m s−1. Phytoplankton bloom occurs when upwelling intensity is ranging between approximately 0.90 m/day – 3.50 m/day at a lag of about 2–3 days. Chl-a, aph_443, adg_443 and POC concentrations show simultaneous increment after upwelling and Chl-a is found to be the major parameter controlling other bio-optical properties of the study region. The waters belong to Case-1 type mainly co-dominated by phytoplankton and CDOM with very low concentrations of suspended particulate matter. Additionally, a noticeable drop in SST ranging between 0.55 °C to 1.50 °C has been observed at the same day of upwelling when upwelling intensity is around 1.5 m/day or above.
... INSAT-3DR (Repeat) with the same payload was launched to 74°E in 2016, with INSAT-3DS (Second Repeat) scheduled for launch in 2022. Comparisons with the iQuam drifter data in the seas around India, revealed mean differences ranging from −0.16 K to −0.20 K, being worse in the Arabian Sea at −0.27 K with greater values during the monsoon season (Tyagi et al., 2018). ...
Sea-surface temperature (SST) is on oceanic variable that is both highly important to a wide range of studies and operational applications, and accurately measureable from spacecraft radiometers. SST has been derived for nearly a decade from the MODIS on Terra. MODIS is a very complex instrument that introduced several innovative developments compared to heritage instruments. For the measurement of SST the MODIS design includes two additional, very narrow spectral bands in the mid-infrared window that have reduced sensitivity to variability in atmospheric water vapor distribution, and thus have the potential for increasing the accuracy of the SST retrievals. Further design innovations include an excellent internal black-body thermal reference target for the accurate calibration of the infrared measurements, and multiple detectors in each spectral band and a dual sided paddle-wheel scan mirror. The added complexity of the instrument came at the cost of the need for additional corrections for instrumental artifacts that would be missing in a simpler device, such as the correction for the reflectivity of the scan mirror that changes with angle of incidence as well as wavelength in the thermal infrared. After several iterations of improved instrumental corrections, the SST accuracies from MODIS are comparable to those of the best spacecraft radiometers, and therefore make a contribution to the long-term SST Climate Data Record. In this presentation we discuss the forms of the algorithms used to derive SSTs from the MODIS on-orbit data and detail the approach to validation with a summary of the current best estimates of MODIS SST retrievals.
Validating remotely sensed sea surface temperature (SST) is a fundamental step in establishing reliable biological/physical models that can be used in different marine applications. Mapping SST using accurate models would assess in understanding critical mechanisms of marine and coastal zones, such as water circulations and biotic activities. This study set out to validate MODIS SSTs with a spatial resolution of 1-km in the Arabian Gulf (24–30° N, 48–57° E) and to assess how well direct comparison of dual matchups and triple collocation analyses perform. For the matchup process, three data sets, MODIS-Aqua, MODIS-Terra, and iQuam, were co-located and extracted for 1-pixel box centered at each actual in situ measurement location with a time difference window restricted to a maximum of ±3 h of the satellite overpass. Over the period July 2002 to May 2020, the MODIS SSTs (N = 3786 triplets) exhibited a slight cool night-time bias compared to iQuam SSTs, with a mean ± SD of −0.36 ± 0.77 °C for Aqua and −0.27 ± 0.83 °C for Terra. Daytime MODIS SST observations (N = 5186 triplets) had a lower negative bias for both Aqua (Bias = −0.052 °C, SD = 0.93 °C) and Terra (Bias = −0.24 °C, SD = 0.90 °C). Using extended triple collocation analysis, the statistical validation of system- and model-based products against in situ-based product indicated the highest ETC-based determination coefficients (ρt,X2 ≥ 0.98) with the lowest error variances (σε2 ≤ 0.32), whereas direct comparison underestimated the determination coefficients and overestimated the error estimates for all MODIS algorithms. The ETC-based error variances for MODIS Aqua/Terra NLSSTs were 0.25/0.19 and 0.26/0.32 in daytime and night-time, respectively. In addition, MODIS-Aqua was relatively more sensitive to the SST signal than MODIS-Terra at night and vice versa as seen in the unbiased signal-to-noise ratios for all observation types.
Sea surface temperature (SST) is an important parameter used to describe the air-sea interaction and the state of marine structures. Atmospheric water vapor has a significant attenuation effect on thermal infrared information, which will reduce the accuracy of SST inversion. However, the effect of atmospheric water vapor on the inversion accuracy is not considered carefully in the existing SST inversion algorithm. In this study, a new method is proposed for retrieving SST from Landsat 8 Thermal Infrared Remote Sensing (TIRS) data based on the variation of atmospheric water vapor content (wvc). First, simulating atmospheric conditions by using moderate resolution atmospheric transmission (MODTRAN) based on atmospheric profiles data (air temperature and pressure). Second, calculating the bright temperature based on the radiation transfer equation and Planck's law. Then, constructing the SST retrieval model of thermal infrared remote sensing based on wvc. Finally, evaluating the accuracy of the proposed model by using simulation data. The root mean square errors (RMSEs) are within 0.5 K, indicating that the accuracy of the model is good in theory. In addition, taking the Zhoushan sea area as the research area, SST is retrieved by Landsat 8 TIRS data. The accuracy of inversion results is evaluated by advanced very high-resolution radiometer (AVHRR) SST products. The bias and RMSE based on the AVHRR SST products are within 1 K and 2 K, respectively. The results show that the accurate SST with high spatial resolution was successfully obtained by using the method. The study is of great significance to the acquisition of marine structural parameters, the exploitation of marine resources, and the monitoring of marine disasters.
