Willmott Index Of Agreement
The first case of study is satellite measurements of the Standardized Difference Vegetation Index (NDVI) obtained from October 1, 2013 to May 31, 2014 on Northwest Africa. The spatial resolution is 1 km and the temporal resolution is a decade (a decade is a period that results from the division of each calendar month into 3 parts, which can take values of 8, 9, 10 or 11 days). The data are obtained from two different instruments on two different satellite platforms: SPOT-VEGETATION and PROBA-V (these are called VT and PV for simplicity). PV data is available through the copernicus Global Land Service Portal24, while VT archive data is provided courtesy of the GFC MARSOP25 project. Although the geometric and spectral characteristics of satellites and data processing chains have been as close as possible, differences between products are still expected because the instruments are not identical. The aim here is to quantify where the time series do not coincide in the region. Since there is no reason to argue that one should be a better reference than the other, a symmetrical match index should be applied to each pair of time series, resulting in values that can be attributed geographically. An important point about the non-cryptic index is that, although it does not take into account any bias, it does not mean that it corresponds to the correlation. The two also do not correspond α in the case of Difference can be appreciated by looking at a cloud of dots and turning it.
This changes their correlation, but thanks to the clean fences, will remain the same. This also leads to positive values for when, which can be interpreted as noise levels in the data. Among the statistical indices, some quantify the difference in the emission of observed or experimental measurement models, while others focus on the correlation between model forecasts and measurements. Essentially, Fox (1981) recommended calculating and reporting the following four types of differences: average error, average absolute error, variance in the distribution of difference, and defects of the root ailment square (or its square – the average quadratic error). These difference-based statistics quantify the output output of the measurement model. Specific indicators are also proposed. Bellocchi et al. (2002) proposed an expert system to calculate a composite indicator of solar radiation performance assessment. They used a correlation coefficient (r), a relative average value error (RRMS), modeling efficiency (EF) and student probability t to form an aggregated form. Confalonieri et al.
(2010) proposed a fuzzy indicator to assess soil water simulation. Jacovides and Kontoyiannis (1995) proposed moderate bias errors (MBE) and square value errors (RMSE) in combination with t statistics as statistical indicators for the evaluation and comparison of evapotranspiration computational models. Among differential and/or statistical ratios, average error (ME), mass error (RMSE), relative error (UC) and correlation coefficient (r) are widely used in different areas – crop growth and yield (Geerts et al. 2009), irrigation planning (Liu et al. 1998), hydrological (Shen et al. 2009), environment (Wagener and Kollat 2007), sun exposure (Rivington et al. 2005), pollution simulation model (Yang et al. 2007), etc. Model efficiency (EF) is used in almost all simulation areas.
The above indices are used both for the evaluation of the individual model and for the comparison of several models (Prasher et al.