ABSTRACT
A framework is proposed for assessing the physical consistency between two terrestrial Essential Climate Variables (ECVs) products retrieved from Earth Observation at global scale. The methodology assessed the level of agreement between the temporal variations of Leaf Area Index (LAI) and Fraction of Absorbed Photosynthetically Active Radiation (FAPAR). The simultaneous changes were classified according to their sign, magnitude and level of confidence, whereby the respective products uncertainties were taken into consideration. A set of proposed agreement metrics were used to identify temporal and spatial biases of non-coherency, non-significance, sensitivity and the overall level of agreement of the temporal changes between two ECVs. We applied the methodology using the Joint Research Center (JRC) Two-stream Inversion Package (TIP) products at 1 km, those provided by the Copernicus Global Land Service (CGLS) based on the SPOT/VGT and Proba-V at 1 km, and the MODIS MCD15A3 at 500 m. In addition, the same analysis was applied with aggregated products at a larger scale over Southern Africa. We found that the CGLS LAI and FAPAR products lacked consistency in their spatial and temporal changes and were severely affected by trends. The MCD15A3 products were characterized by the highest number of non-coherent changes between the two ECVs but temporal inconsistencies were mainly located over the eastern hemisphere. The JRC-TIP products were highly consistent. The results showed the advantages of physically-based retrieval algorithms, in both JRC-TIP and MODIS products, and indicated also that, except for MODIS over forests, aggregated products using an uncertainty-based weighted average led to higher agreement between the ECVs changes.
ABSTRACT
This paper proposes a benchmarking method for assessing the level of spatio-temporal variability of Essential Climate Variable (ECV) products against a reference taking into account acceptance criteria in terms of intensity and physical distance tolerances. This is based on a modified version of the gamma index that could be suitable for fitness-for-purpose assessment given that one can choose various criteria depending on applications. The method is first presented and then applied to both land and atmospheric ECVs. The terrestrial analysis concerns the global surface albedo, using monthly white-sky surface albedo in the visible, near-infrared and shortwave broadband spectral ranges at a spatial resolution of 0.05° using three sources of products. The latter study is conducted using monthly aerosol optical depth (AOD) products at 550 nm at a spatial resolution of 1° with four different datasets at the global scale. The analysis shows how the values of the gamma criteria impact the spatial and temporal results. As an example, if the Global Climate Observing System (GCOS) actual target measurements uncertainty is used as an acceptance criteria for the intensity tolerance the results show that: 1) the seasonal agreement for the surface albedo products varies over 20% to 40% of the terrestrial surface in the shortwave and near-infrared broadband and from 10% to 30% in the visible one and 2) the three aerosols optical depth products agree with the reference one for over 50% of the land surface only when the tolerance distance term is at 224km.
ABSTRACT
The purpose of this work is to investigate the capability of the FLUKA Monte Carlo (MC) code to simulate the Elekta Leksell Gamma Knife Perfexion (LGK-PFX) and reproduce the Treatment Planning System (TPS) Leksell GammaPlan version 8.2 (LGP) dose calculations for the case of a water equivalent phantom target. Thanks to the collaboration with Elekta Instruments AB, the collimation system geometry, the source positions and all the involved material have been simulated in detail. The relative linear dose distribution along the three coordinate axes, for each collimator size, and the Relative Output Factors (ROF) have been investigated. The simulation has been validated comparing simulated linear dose profiles with measurements performed with EBT radiochromic films. The acceptance criterion between experimental data and FLUKA results is based on the gamma index (GI) method. The FLUKA MC calculation for the ROF provided the values of 0.920 for the 8 mm collimators and 0.800 for the 4 mm collimators. These values are in good agreement with the Elekta reference data of 0.924 and 0.805 respectively. The percentage difference between calculated and reference values for the ROF is under 1% and within the FLUKA uncertainty. Also the simulated relative dose profiles show a good agreement with the LGP calculation expressed by means of the gamma index method. This established accuracy proves that FLUKA is a suitable and powerful tool in order to reproduce successfully the LGP calculations for the homogeneous media.
