Photosynthetically Active Radiation (PAR)

Last update: Jan. 2019

    Objective - Characteristics of the service - Steps of developments - Bibliography


Objective PAR service

  • Evaluate the performance of the two methods of Wandji Nyamsi et al. 2015 (respectively named "weighted_kato" and the most performing one "discretized_Kato") against three other methods: Udo et Aro (1999), Jacovides et al. (2004), Sceicz (1974).
  • Develop an operational service for discretized_Kato in all-sky conditions in the geographical coverage of the Meteosat Second Generation satellite since Feb. 2004 onwards, from 1 min to 1 month time step. This service will deliver time series of spectral radiation values in the spectral band interval required by the user.

Framework - collaboration

University of East Anglia. Comparison of the PAR data provided by the 5 methods against the measurements of:

  • Aberystwyth University (30 min means, middle of interval UTC)
  • Abbotts Hall (idem)
  • Cartmels Sands (idem)
  • and the updated measurements from 4 soft fruit farms from Apr. 2018

Click to magnify=>


To do: give the validation results for the three sites with long-term measurements + upload poster EMS 2018

Steps of developments

(reverse chronological order)

29 Nov. 2017: rdv au CIEMAT. Discussion with the team about spectral developments

Oct. 2017: McClear v2 to McClear v3 => Stéphane is checking if ok. Correction of compute_gc_mcclear to correct the use of E0 (broadband instead of Kato)

Sept. 2017: McClear in real time

Aug. 2017: Self validation and first delivery of spectral radiation datasets to a user. Validation against in situ measurements is not yet carried out.

July: Bug in McClear v2 (change of atmospheric profile during the day generating sharp steps of radiation values). This has been corrected during the maintenance operation that occurred on 31st July 2017 (information not available in the release not as it impacted only the spectral data which is not an operational service).

Spring 2017: First developments of a precursor of service in Matlab following the work of William

2016 and before: research carried out by William Wandji, a PhD student from MINES ParisTech


Monthly and daily PAR from 2000-2016 :
Result from a radiation model transfer + neural networks with forcings from MODIS atmospheric products

Huete et al (2002) explique que l'indice EVI est plus adapté pour décrire l'activité photosynthétique des forêts tropicales (high-biomass forests) que le NDVI. (See Wikipedia for EVI - the values of coefficients for MODIS-EVI algorithm are; L=1, C1 = 6, C2 = 7.5, and G (gain factor) = 2.5)

PAR sensor, PSQ1 INRA Bornet June 2018: "A standardised PAR spectral response in the visible light range between 400nm and 700nm wavelength was defined by McCree (1972) such that each photon within this region is equally absorbed. ‘Blue' photons of shorter wavelength (higher frequency) have more energy than ‘Red' photons of longer wavelength. The amount of PAR is commonly expressed as Photosynthetic Photon Flux Density (PPFD) with a unit of µmol/m²·s."

Need for an online tool to convert µmol/m2/s into W/m2 or Wh/m2.

[le stress des plantes soumises à des grandes variabilités temporelles de l'éclairement solaire]

Proposition of division of the visible spectral by Dario: (not compliant with  ISO 21348 Definitions of Solar Irradiance Spectral Categories)

  • blu-violette, (400-490 nm),
  • green (490-560 nm),
  • yellow (560-590 nm);
  • red-orange (590-700 nm)   

Top of page