Gap-filling and quality control in biometeorological modelling: overview of best practices

Biometeorological models simulating the dynamics and spread of insect species coupled with spatial decision support systems (SDSS) are becoming an important component of human and animal disease risk assessment and mitigation planning. To accurately simulate the dynamics of insect species and the anticipated spread of vector-borne disease (VBD), complex mathematical models are being developed which require precise and continuous micrometeorological forcing representative of the vector habitat, provided in near-real time. For this reason, reanalysis product such as ERA5-Land are not always suitable, and environmental wireless sensor networks (WSN) are often employed.Regardless of the purpose of the observations, the quality of the recorded data is determined by the accuracy and completeness of the measurements, the spatial and temporal representativeness of the records, and the representativeness with respect to the goal of the observation. With the increasing use of autonomous environmental sensors employing different data-storage, transmission, and communication protocols, particularly in low-power/low-cost applications, the risk of data loss and measurement error grows, emphasising the importance of properly implementing quality planning and assurance.In this paper we provide an outline of quality control (QC) and gap-filling methods for dealing with spatial and temporal gaps in meteorological measurements for different classes of biometeorological applications with a focus on: (i) mathematical vector population dynamics models; (ii) machine learning vector distribution models; and (iii) mechanistic vector distribution models.Establishing a comprehensive quality assurance (QA) and gap-filling protocol for micrometeorological data in the field of biometeorological modelling will improve the understanding of the types of errors and limitations expected during analysis, as well as provide an overview of best-practices for a broader interdisciplinary audience.