Snow reflectance
factor brdf.pngThis is hard test sampling. The spatial resolution is set to 10 mts, the spectral resolution is set to 2 nm.
A year of TOA snow
Pixel saturation map.
Daily heatmap spectrum in scene with clouds, snow and terrain
One year of column water wapor.
One year of aerosol content.
TOA, averaging RGB using full spectrum
Lambertian, clouds and shadows (mountains and clouds)
The scope of this work is to provide a tool to evaluate the dynamic range of scenes in a large variety of contexts these contexts includes:
A variation of the CHIME Phase A/B1, has been prepared to provide comparison of the reference level signal modeled by a simple exponential model to hold, max signal, reference singal and min signal.
The scene selected is a region of Europe in the Alpes region. This region contains vegetation.
The program has been modified to provide synthetic snow in a range of elevations and can be selected depending on the day or provide a fixed value.
The snow reflectance values has been taken from the SCOPE RTM, and are provide with the exact values for all the demostrator.
Snow reflectance
factor brdf.png
Also the component of full specular reflection is presented to evaluate the sun position with the sensor point of view.
BRDF factor (sun/sensor)
synthetic_snow_reflectivity.png
Not required by the model but still presented the specular component of the terrain is presented.
Specular component (sun/sensor)
surface_specular_angle.png
Ths results presented in only snow shows different outputs of the simulation. A fictional terrain including elevation at sea level, to stress the dynaic range is show in here, where clouds, snow, real terrain and falsed elevation is also presented.
Pixel saturation is declared if only one band is saturated, so a study of how many bands are saturated in the pixel has been also performed.
As major result we show there is a large influence of the coupling with the selected terrain model with the reflectance. This provides a 10% of pixels with saturation.
Synthetic snow mask
synthetic_snow_mask.png
In terms of number of channels, for each ‘year day’.
Bands affected by day, each day is a
histogram of number of bands affected saturation.png. The
simulation starts in July, so maximum of illumination is provided at the
beginning and the end of the simulation.
The SGM program has been desinged to support realistic maps, and provide combination of real DEM, scenes, vegetation RTM, synthetic scenes and more.
Sources of realistics scenes can be provided by feeding with a map of pixels with the reflectivity at several resolution, the program will interpolate the reflectivity to better adjust to the required resolution, allowing the use of historic data.
A large number of features to provide automatic scenes, has been implemented, and one of the best characteristics is the speed.
For this sensor scenes can be provided in less than 10’ using 8 CPU cores of modern intel processors, also is able to scale linearly to a large number of processors. The idea about paralelization is provided by using a non blocking system, combining non blocking consumer producer model.
This model allows calculation of realistic terrain kernels with no major impact of the memory. This means that to provide a scene of 175Gb is not required more than 3 Gb of machine, in this 8 Cores. Memory is relaxed as soon as it is written.
This perfomance gain allows large studies. For this work 6 one year simulations under several conditions has been produced, meaning that 6 orbit’s range from Nowergian to South Africa, can be performed, allowing a large number of sun/atmosphere studies.
Compiting model
computing_model.png
This scenes starts at 20 Jul, are a full year of TOA in the same region with daily different atmosphere conditions. Sun and all elevation is activated.
For more information, please contact us at: