Decision Support Systems (DSSs) are used in precision agriculture to provide feedback to a variety of stakeholders, including farmers, advisers, researchers and policymakers. However, increments in the amount of data might lead to data quality issues, and as these applications scale into big, real-time monitoring systems the problem gets even more challenging. Visualisation is a powerful technique used in these systems that provides an indispensable step in assisting end-users to understand and interpret the data. In this paper, we present a systematic review to synthesise literature related to the use of visualisation techniques in the domain of agriculture. The search identified 61 eligible articles, from which we established end-users, visualisation techniques and data collection methods across different application domains. We found visualisation techniques used in various areas of agriculture, including viticulture, dairy farming, wheat production and irrigation management. Our results show that the majority of DSSs utilise maps, together with satellite imagery, as the central visualisation. Also, we observed that there is an excellent opportunity for dashboards to enable end-users with better interaction support to understand the uncertainty of data. Based on this analysis, we provide design guidelines towards the implementation of more interactive and visual DSSs.

No abstract

Validating coarse scale remote sensing soil moisture products requires a comparison of gridded data to point-like ground measurements. The necessary aggregation of in situ measurements to the footprint scale of a satellite sensor (>100 km²) introduces uncertainties in the validation of the satellite soil moisture product. Observed differences between the satellite product and in situ data are therefore partly attributable to these aggregation uncertainties. The present paper investigates different approaches to disentangle the error of the satellite product from the uncertainties associated to the up-scaling of the reference data. A novel approach is proposed, which allows for the quantification of the remote sensing soil moisture error using a temporally adaptive technique. It is shown that the point-to-area sampling error can be estimated within 0.0084 [m³/m³]

Abstract.Validating coarse scale remote sensing soil moisture products requires a comparison of gridded data to pointlike ground measurements. The necessary aggregation of in situ measurements to the footprint scale of a satellite sensor (>100 km 2 ) introduces uncertainties in the validation of the satellite soil moisture product. Observed differences between the satellite product and in situ data are therefore partly attributable to these aggregation uncertainties. The present paper investigates different approaches to disentangle the error of the satellite product from the uncertainties associated to the up-scaling of the reference data. A novel approach is proposed, which allows for the quantification of the remote sensing soil moisture error using a temporally adaptive technique. It is shown that the point-to-area sampling error can be estimated within 0.0084 [m 3 /m 3 ].

No abstract