The efficient detection of cloud scenes by Radiance Enhancement (RE) and with their impact on earth global energy budget due to Short Wave upwelling Radiative Flux (SWupRF) within NIR spectral

range of space-orbiting Argus1000 micro-spectrometerRehan Siddiqui1, Naif Al Salem1, Rajinder Jagpal1, Brendan Quine1,2

1. York University, Toronto, Canada 2. Thoth Technology Inc., CanadaAbstract

The Radiance Enhancement (RE) and integrated absorption technique isapplied to develop a clouds model by enhancement in radiance due to cloud andno cloud phenomena. The Shortwave upwelling Radiative Flux (SWupRF)W/m

2

within NIR of Argus 1000 for O2, H2O, CO2 and CH4 has also been quantifies.This new model is used to estimate the magnitude variation for RE andSWupRF over spectral range of 1100nm to 1700nm by varying surface altitude,mixing ratios and surface reflectivity. In this work we employ satellite realobservation of space orbiting Argus 1000 along with line by line GENSPECTradiative transfer model for the efficient detection of clouds with their impact onsurface energy budget due to SWupRF effects. We calculate and compare boththe synthetic and real measured observed data set of different week per pass ofArgus flight. Results are found to be comparable for both the approaches, afterallowing for the differences with the real and synthetic technique. Themethodology based on RE and (SWupRF)W/m

2 of the space spectral data can bepromising for the instant and reliable detection of the cloud scenes with theirimpact on earth climate.

Introduction

• Clouds are central phenomena to provide a link between the two key energyexchange processes that determine the earth climate, namely solar andterrestrial radiance exchanges and water exchanges.

• Clouds play a very important role in controlling earth’s climate and directlychange the cooling or heating of the surface below.

• Identification of clouds by radiance enhancement( RE) and shortwaveupwelling radiative flux SWupRF (W/m2) in remote sensing dataset are a keyissue particularly in the case of sensors working in the visible and near infraredrange of electromagnetic spectrum due to severe absorption of cloudconstituent’s.

• The incoming solar radiation is attenuated as it penetrates the atmosphere,reflects from the surface and travels back to space.

• In real atmosphere, the attenuation includes the molecular (Rayleigh)scattering, absorption by CO2, CH4, CO and water vapor in a form of clouds.The most popular greenhouse gas H2O plays a very important part for cloudsanalysis.

• Clouds with radiation energy usually play a very important role throughabsorption and scattering of photonic radiance within different atmosphericlayers specially water vapor, carbon dioxide, oxygen and aerosols.

• Clouds effect the path of photons through the atmosphere and thereforechange the interpretation of the depth of an absorption band. In cloud retrievalsfrom a satellite, it is very important to have a good estimate of the surfacealbedo. The reason is that the cloud detection is usually performed bycomparing the measured reflectance with expected reflectance from cloudscene

• Measurements acquired from the space orbiting Argus 1000 provide vitalinformation related to long term changes in atmospheric composition andclouds scenes in terms of their radiative effects.

• In this study a new RE & SWupRF techniques are applied to space baseddataset of Argus1000 micro-spectrometer along with GENSPECT line-by-lineradiative transfer model to detect cloud scene.

Argus 1000 spectrometer

Conclusion

• The Argus 1000 is a micro-spectrometer, launched on aboard on the 28th of April 2008 onthe Canadian Advanced Nano space eXperiment 2 (CANX-2).

• Argus 1000 has 136 channels in the near infrared spectrum 1100 – 1700 nm with a spectralresolution of 6 nm and an instantaneous spatial resolution of 1.4 km at 640 km orbit.

• Monitor radiation absorption by major atmospheric trace greenhouse gases includingOxygen (O2), Water Vapor (H2O), Carbon Dioxide (CO2) and Methane (CH4).

• The Argus 1000 spectrometer is capable of monitoring ground-based sources and sinks ofanthropogenic pollution. The instrument was designed to take nadir observations of reflectedsunlight from Earth's surface and atmosphere. The nadir viewing geometry of Argus is ofparticular utility as this observation mode provides the highest spatial resolution on the brightland surfaces and returns more useable soundings in regions that are partially cloudy orhave significant surface topography.

• York University has been working with the spacecraft operations at the University of TorontoInstitute for Aerospace Studies Space Flight Laboratory (UTIAS-SFL) to process the sciencedata and the related spacecraft telemetry, principally measurement timing and spacecraftpointing, in order to properly interpret the Argus Observations.

GENSPECT line-by-line radiative transfer code• The GENSPECT is a line-by-line radiative transfer algorithm for absorption, emission, and

transmission for a wide range of atmospheric gases.• GENSPECT can compute synthetic spectra for comparison with data collected by Earth observing

instruments. The radiative transfer characteristics of gases have been extensively analyzed and arecommonly modeled by examination of the energy transitions between different vibrational states of agas molecule.

• The loader input parameters includes gas file, specific gas name, run time, wavenumber, airbroadening, self-broadening, temperature coefficients etc.

• Each atmospheric species has a unique absorption cross section spectrum, which is used widely toidentify the species in atmospheric spectral measurements.

• GENSPECT has been used to compute synthetic spectra for comparison with data collected by Earthobserving instruments.

Methodology (RE vs. SWupRF)• An integrated technique have been developed to calculate the Short Wave upwelling Radiative Flux (SWupRF) in term of

W/m2 both for Synthetic and Argus observed spectra by using GENSPECT line by line radiative transfer model.• The detection of cloud scene can be found by the difference of the ratio of the observed data with simulated data for the

selected week/pass of Argus flight with single or multiple scan (RE technique).• SHARCNET (SAW & ORCA) clusters are used to run many GENSPECT-Synthetic jobs both in series and parallel mode.• ‘Lsqnonlin’ non-linear optimization technique is applied on different GENSPECT jobs along with Argus dataset.• The relationship of RE and SWupRF are as follows:

RE & SWupRF results

•McNally, A.P. and Watts, P.D., 2003. A cloud detection algorithm for high‐spectral‐resolution infrared sounders. Quarterly Journal of the Royal Meteorological Society, 129(595), pp.3411-3423.•Cutillo, L., Amato, U., Antoniadis, A., Cuomo, V. and Serio, C., 2004. Cloud detection from multispectral satellite images. In Proceedings from IEEE Gold Conference, University Parthenope, Naples, Italy.•Song, X., Liu, Z. and Zhao, Y., 2004, September. Cloud detection and analysis of MODIS image. In Geoscience and Remote Sensing Symposium, 2004. IGARSS'04. Proceedings. 2004 IEEE International (Vol. 4, pp. 2764-2767). IEEE.•Minnis, P., Sun-Mack, S., Chen, Y., Yi, H., Huang, J., Nguyen, L. and Khaiyer, M.M., 2005, October. Detection and retrieval of multi-layered cloud properties using satellite data. In Remote Sensing (pp. 597909-597909). International Society for Optics and Photonics.•Chylek, P., Robinson, S., Dubey, M.K., King, M.D., Fu, Q. and Clodius, W.B., 2006. Comparison of near‐infrared and thermal infrared cloud phase detections. Journal of Geophysical Research: Atmospheres, 111(D20).•Tang, B., Li, Z.L. and Zhang, R., 2006. A direct method for estimating net surface shortwave radiation from MODIS data. Remote Sensing of Environment, 103(1), pp.115-126.•Cheng, T., Gu, X., Chen, L., Yu, T. and Tian, G., 2007, July. Cloud detection based on the spectral, multi-angular, and polarized characteristics of cloud. In 2007 IEEE International Geoscience and Remote Sensing Symposium (pp. 3321-3324). Ieee.•Li, W. and Li, D., 2008, October. Cloud detection in MODIS data based on spectrum analysis and snake model. In Geoinformatics 2008 and Joint Conference on GIS and Built Environment: Classification of Remote Sensing Images (pp. 714710-714710). International Society for Optics and Photonics.•Li, D., Dong, X., Liu, L. and Xiang, D., 2008, January. A new cloud detection algorithm for FY-2C images over China. In Proceedings of the 1st international conference on Forensic applications and techniques in telecommunications, information, and multimedia and workshop (p. 52). ICST (Institute for Computer Sciences, Social-Informatics andTelecommunications Engineering).•Krezel, A. and Paszkuta, M., 2011. Automatic detection of cloud cover over the Baltic Sea. Journal of Atmospheric and Oceanic Technology, 28(9), pp.1117-1128.•Ghosh, R.R., Ali, M.S., Hena, A. and Rahman, H., 2012. A Simple Cloud Detection Algorithm Using NOAA-AVHRR Satellite Data. International Journal of Scientific & Engineering Research, 3(6).•Kazantzidis, A., Tzoumanikas, P., Bais, A.F., Fotopoulos, S. and Economou, G., 2012. Cloud detection and classification with the use of whole-sky ground-based images. Atmospheric Research, 113, pp.80-88.•de Leeuw, G., Kokhanovsky, A.A. and Cermak, J., 2012. Remote sensing of aerosols and clouds: Techniques and applications (editorial to special issue in Atmospheric Research). Atmospheric Research, 113, pp.40-42.•Escrig, H., Batlles, F.J., Alonso, J., Baena, F.M., Bosch, J.L., Salbidegoitia, I.B. and Burgaleta, J.I., 2013. Cloud detection, classification and motion estimation using geostationary satellite imagery for cloud cover forecast. Energy, 55, pp.853-859.•Tang, H., Yu, K., Hagolle, O., Jiang, K., Geng, X. and Zhao, Y., 2013. A cloud detection method based on a time series of MODIS surface reflectance images. International Journal of Digital Earth, 6(sup1), pp.157-171.•Yordanov, G.H., Midtgård, O.M., Saetre, T.O., Nielsen, H.K. and Norum, L.E., 2013. Overirradiance (cloud enhancement) events at high latitudes. IEEE Journal of Photovoltaics, 3(1), pp.271-277.•Fisher, A., 2014. Cloud and cloud-shadow detection in SPOT5 HRG imagery with automated morphological feature extraction. Remote Sensing, 6(1), pp.776-800.•Wang, T., Yan, G., Shi, J., Mu, X., Chen, L., Ren, H., Jiao, Z. and Zhao, J., 2014, July. Topographic correction of retrieved surface shortwave radiative fluxes from space under clear-sky conditions. In 2014 IEEE Geoscience and Remote Sensing Symposium (pp. 1813-1816). IEEE.•Mateos, D., Antón, M., Valenzuela, A., Cazorla, A., Olmo, F.J. and Alados-Arboledas, L., 2014. Efficiency of clouds on shortwave radiation using experimental data. Applied Energy, 113, pp.1216-1219.•Illingworth, A.J., Barker, H.W., Beljaars, A., Ceccaldi, M., Chepfer, H., Clerbaux, N., Cole, J., Delanoë, J., Domenech, C., Donovan, D.P. and Fukuda, S., 2015. The EarthCARE satellite: The next step forward in global measurements of clouds, aerosols, precipitation, and radiation. Bulletin of the American Meteorological Society, 96(8), pp.1311-1332.•Guo, F., Shen, X., Zou, L., Ren, Y., Qin, Y., Wang, X. and Wu, J., 2015. Cloud Detection Method Based on Spectral Area Ratios in MODIS Data. Canadian Journal of Remote Sensing, 41(6), pp.561-576.•Siddiqui, R., Jagpal, R., Salem, N.A. and Quine, B.M., 2015. Classification of cloud scenes by Argus spectral data. International Journal of Space Science and Engineering, 3(4), pp.295-311.•Xu, D., Auligné, T. and Huang, X.Y., 2015. A validation of the multivariate and minimum residual method for cloud retrieval using radiance from multiple satellites. Advances in Atmospheric Sciences, 32(3), pp.349-362.•Menaka, E., Kumar, S.S. and Bharathi, M., 2015. Cloud removal using efficient cloud detection and removal algorithm for high–resolution satellite imagery. International Journal of Computer Applications in Technology, 51(1), pp.54-61.•Jagpal, R.K., Quine, B.M., Chesser, H., Abrarov, S.M. and Lee, R., 2010. Calibration and in-orbit performance of the Argus 1000 spectrometer-the Canadian pollution monitor. Journal of Applied Remote Sensing, 4(1), pp.049501-049501.•Quine, B.M. and Drummond, J.R., 2002. GENSPECT: a line-by-line code with selectable interpolation error tolerance. Journal of Quantitative Spectroscopy and Radiative Transfer, 74(2), pp.147-165.•Chylek, P. and Borel, C., 2004. Mixed phase cloud water/ice structure from high spatial resolution satellite data. Geophysical research letters, 31(14).•Jagpal, R.K., 2011. Calibration and Validation of Argus 1000 Spectrometer–A Canadian Pollution Monitor (Doctoral dissertation, YORK UNIVERSITY, TORONTO).•Jagpal, R.K., Quine, B.M., Chesser, H., Abrarov, S.M. and Lee, R., 2010. Calibration and in-orbit performance of the Argus 1000 spectrometer-the Canadian pollution monitor. Journal of Applied Remote Sensing, 4(1), pp.049501-049501.•Chesser, H., Lee, R., Benari, G., Jagpal, R., Lam, K. and Quine, B., 2012. Geolocation of Argus flight data. IEEE Transactions on Geoscience and Remote Sensing, 50(2), pp.357-361.

Key References

Fig. 5. Argus 1000 qualification model(image courtesy of Thoth Technologies, Inc.).

Fig. 6. Argus Field-of-view

Fig. 1. Earth and its atmosphere maintain an energybalance by either absorbing incoming radiation orreflecting it energy back into space.

Fig. 2. The short wave and long wave component of theenergy budget.

Fig. 4. CanX-2-Argus1000 observation geometry for the detectionof different atmospheric features.

Fig. 3. Canadian Advanced Nanosatellite eXperiment-2(CanX-2) with Argus 1000, an atmospheric-spectrometerwithin the spectral range of 900-1700 nm.

Fig. 11. GENSPECT-Synthetic model from 0% to 100% of H2Oconcentration (by using US atmospheric model) .

Fig. 13. GENSPECT-Synthetic model with Argus 1000 – Solarzenith & sun angle.

Fig. 14. GENSPECT-Synthetic model with Argus 1000 – Albedoand Altitudes variations.

Table I : Typical Argus week per pass - Parameters for sampling of calculating RE & SWupRF.

Fig. 20. Argus observed spectra of week 14 pass 52 (March 4th , 2010) over Toronto/Kitcisakik (Canada) with observation numbers 22 to 125.

Fig. 23. Argus week 75 pass 43 (August 14th , 2013) withSWupRFobs (0.20- 2.22 W/m2) and variability factor 11.0

Fig. 22. Argus observed spectra of week 75 pass 43 (August 14th ,2013) over North pacific ocean with observation numbers 19 to 65.

Fig. 18. Argus observed spectra of week 8 pass 61 (October 30th ,2009) over Arabian sea with observation numbers 64 to 238.

Week No. _Pass No. Date Observation Numbers (OBS)

Packet Length

Observations number with satellite Sun

angle, Nadir angle, Lat. & Long.

Location

Week08_Pass61 2009October30 32-121 276

OBS32: Sat. nadir angle = 6.2240 Sat. sun angle = 35.4821 Lat. = 8.6388, Long. = 61.2715 OBS121: Sat. nadir angle = 4.9744

Sat. sun angle = 33.0354 Lat. = 3.0414, Long. = 60.2615

Arabian Sea &

Seychelles

Week09_Pass36 2009November04 22-45 69

OBS22: Sat. nadir angle = 5.0968 Sat. sun angle = 35.5000

Lat. = 6.8350, Long. = -35.4025 OBS45: Sat. nadir angle = 5.4026 Sat. sun angle = 33.9916 Lat. = 3.7837, Long. = -36.0342

North Atlantic Ocean

Week14_Pass52 2010March04 22-125 125

OBS22: Sat. nadir angle = 24.7152 Sat. sun angle = 58.1136 Lat. = 47.3740, Long. = -77.7286 OBS125: Sat nadir angle = 21.8777 Sat. sun angle = 54.5434

Lat. = 42.2893, Long. = -79.9453

Toronto/ Kitcisakik (Canada)

Week14_Pass54 2010March04 14-123 123

OBS14: Sat. nadir angle = 4.1527 Sat. sun angle = 62.7967 Lat. = 53.1442, Long. = -124.6127 OBS123: Sat nadir angle = 17.2101 Sat. sun angle = 57.9201 Lat. = 47.0893, Long. = -126.6271

Vancouver (Canada)

Week17_Pass42 2010April28 14-144 144

OBS14: Sat. nadir angle = 18.2065 Sat. sun angle = 51.0290 Lat. = 64.8690, Long. = 177.4778 OBS144: Sat nadir angle = 1.2381 Sat. sun angle = 43.9990

Lat. = 55.5832, Long. = 172.3671

Magadan (Russia)

Week30_Pass46 2010December 16 35-460 460

OBS35: Sat. nadir angle = 2.6513 Sat. sun angle = 32.3060 Lat. = -26.7341, Long. = 43.2740 OBS460: Sat nadir angle = 2.2886 Sat. sun angle = 45.0506

Lat. = -55.4500, Long. = 33.7905

Indian Ocean

Week41_Pass27 2011September08 14-198 198

OBS14: Sat. nadir angle = 4.5387 Sat. sun angle = 56.8632 Lat. = -37.5321, Long. = -72.0149 OBS198: Sat nadir angle = 3.0990 Sat. sun angle = 67.1288 Lat. = -50.2686, Long. = -76.2329

Patagonia (South America)

Week75_Pass43 2013August14 19-65 65

OBS19: Sat. nadir angle = 1.8888 Sat. sun angle = 38.6809 Lat. = 31.7593, Long. = 148.2136 OBS65: Sat nadir angle = 1.6877 Sat. sun angle = 38.1652

Lat. = 19.9247, Long. = 145.5481

North Pacific Ocean

Types of parameter Significance values and ranges

Mixing Ratios of gases

refmod 95_ O2.mxr, refmod 95_ CO2.mxr, refmod 95_

CH4.mxr, refmod 95_ H2O.mxr (1976 U.S. Standard

Atmospheric Model)

Gases in % O2 (0-100) , CO2 (1-200), CH4 (1-100), H2O (1 to 100)

Height from surface to top

of clouds2km to 50 km

Surface Type Lambertian

Satellite sun angle* Argus geo location (Ave. of obs no.)

Satellite nadir angle* Argus geo location (Ave. of obs no.)

Reflectivity

0.01 to 0.3 (over oceans)

0.3 (over generic vegetation and bare soil)

0.1 to 0.9 (over snow, clouds, and ice)

Scattering Type Rayleigh

Table II: Input parameters for GENSPECT Synthetic calculations.

Fig. 12. GENSPECT-Synthetic model from 0% to 100% of CO2concentration (by using US atmospheric model) .

1

[ ] [ ]1[ ]

Ni i

ij i

j jN j

OBS SYNRE SYN=

− =

∑

Fig. 7. Argus spectra- radiance vs wavelength of weeks per passes with selected observation numbers (week08/09/14, pass61/36/52/54, obs32/45/22/14/123).

Fig. 8. Argus spectra- wavelength vs radiance of different weeks per passes with selected observation numbers week17/30/41/75, pass42/46/27/43, obs14/144/35/19/65).

1100 1200 1300 1400 1500 1600 17000

5

10

15

20

25

30

35

40

45

50

Wavelength [nm]

Rad

ianc

e [W

(m- 2

)sr- 1

(cm

-1)-1

]

Radiance vs Wavelength of different weeks/passes/observations

W08P61O32

W09P36O45

W09P36O22W14P52O22

W14W54O14

W14P54O123

Fig. 9. Argus spectra- wavelength vs radiance of different weeks per passes with selected observation numbers week17/30/41/75, pass42/46/27/43, obs14/144/35/19/65).

Fig. 10. Argus spectra- wavelength vs radiance of different weeks per passes with selected observation numbers week17/30/41/75, pass42/46/27/43, obs14/144/35/19/65).

Fig. 19. Argus week 8 pass 61 (October 30th , 2009) withSWupRFobs (0.42- 2.35 W/m2) and variability factor 5.6

Fig. 21. Argus week 14 pass 52 (March 4th , 2010) withSWupRFobs (0.58- 2.22 W/m2) and variability factor 3.8

Fig. 15. GENSPECT from 0% to 35% ofH2O concentration, albedo = 0.3 showSWupRFsyn range from 0.4725 [0.3950 to0.5500] W/m2

Fig. 16. GENSPECT from 0% to 35% ofH2O concentration, albedo 0.9 showSWupRFsyn range from 1.340 [1.030 to1.650] W/m2

Fig. 17. GENSPECT from 0% to 200% ofCO2 concentration, show SWupRFsyn rangefrom 0.3430 [0.3407 to 0.3454] W/m2

Fig. 24. Histogram of the selected Argusweeks/passes/observations with max = 2.30 w/m2, min = 0.194W/m2, Ave. = 0.84 W/m2 & Sigma = 0.44 (approximately).

Fig. 27. Scattered plot between RE and SWupRF of Argus data set.

-5 0 5 10 15 20 25 30 350

5

10

15

20

25

30

35

40

Combined Radiance Enhancement(Total)

Freque

ncy

RE(max)=30.4114 (Week 8, Pass 61, Obs 116), RE(min) = -1.95594 (Week 75, Pass 43, Obs 37)

Probability of cloud scenes

Fig. 25. Histogram of the subsequent probability of cloud and nocloud scenes.

Fig. 26 (a), (b), (c) & 9d). Scattered plots of O2, H2O, CO2 & CH4 between RE and SWupRF of Argus data set for the efficient detection ofcloud and no cloud scenes.