In 2009, Cabidoche et al proposed a model to predict chlordecone contamination durations in French Antilles soils. That model, called WISORCH and widely accepted by the scientific community, indicated durations of several decades to several centuries according to the type of soil under consideration. Recently, Compte et al revisited and modified WISORCH to again study chlordecone evolution in Antilles soils. Their results, published in 2022 in Environmental Pollution, indicated a chlordecone half-life of only five years whatever the type of soil.

That half-life value indicates a halving of chlordecone contamination every five years and thus the potential for a concentration under detectable levels by horizon 2050 to 2070, i.e., a much faster end to agricultural soil contamination than was previously thought. If confirmed, the results of Compte et al would enable novel perspectives for the management of contamination in agriculture. Indeed, a useable estimation of chlordecone half-life is among the primordial data needed to predict and manage the evolution of this historical pollutant.

However, researchers from the CEA, Inrae, IRD, BRGM and the universities of Évry–Paris-Saclay, Strasbourg, Lorraine and Mayotte have questioned this report of a chlordecone half-life of only five years, publishing their critical analysis in the journal Environmental Pollution.

Under the direction of Pierre-Loïc Saaidi (UMR8030 Genomics Metabolics/Genoscope/CEA-Jacob) and building upon the observation that a five-year half-life is inconsistent with the level of pollution observed today, 30 years after the banning of chlordecone, the team reexamined the sampling protocols, analytical methods, employed data, applied hypotheses and modeling approaches implemented in Compte et al's 2022 study.

Their work revealed several scientific biases in Comte et al's work, fragilizing the conclusions of this latter. To support their arguments, Saaidi's team took the estimations of the amounts of chlordecone used in Antilles agriculture between 1972 and 1993, and simulated the evolution of their concentrations in the soils using Comte et al's model and presuming their five-year chlordecone half-life. The chlordecone concentrations predicted by that simulation for the period 2000–2020 were largely inferior to those actually found in the soils over that same period.

The researchers thus concluded that a hypothesis of long-term contamination, largely superior to a five-year half-life, should remain under consideration by the scientific community. In addition to the analysis, the team provided recommendations to improve the reliability of data that may be used in future predictive models of chlordecone evolution in the soils of the French West Indies.