The IRIDA project (Innovative Remote and ground sensors, data and tools Into a Decision support system for Agricultural Water Management) startet in 2016 and is a EU/JPI WaterWorks 2014 Cofunded call. Partners in the project were from Spain, Italy, Romania and Norway.The goals and methods differed between partners in the South and Norway. The issue in Southern Europe was optimal use of scarce water resources in agriculture. NIBIO on the other hand had its focus on improving the land-use specific hydrological models which are important for 1) predicting flash floods, 2) serve as a base for modelling erosion, nutrient transport and water quality and 3) estimating reliable soil moisture content at plot/field scale, important from a soil protection/- management aspect (timing of tillage operations and sowing and harvesting). Information from ground and air-borne (satellite and drone) measurements were successfully combined with mathematical modelling. During 2017 and 2018 many field measurements of crop height, crop cover and soil moisture were carried out in Gryteland an field scale catchment, nested within the Skuterud. In addition, multispectral images were obtained from a drone and satellite. A challenge has been the very large spatial variation in soil moisture content and crop development. At the same time was cloud formation often a problem. Based on these findings satellite imagery based on Synthetic Aperture Data (SAR, Sentinel 1) was used to quantify soil moisture content. This was also considered suitable for use in a DSS system because of its independence of weather conditions such as cloud formation and as an improvement in modelling the soil moisture content.
A problem with applying hydrological models has been the correct, true simulation of the soil moisture content at plot/field-scale using models. Many of the models (DRAINMOD, SWAT, HBV, PERSiST) applied in simulating runoff from the Skuterud catchment were calibrated/validated against measured runoff, nutrient and soil loss at the outlet. But within the catchment are different soil types and land uses and as such these models were not able to provide detailed information about the soil moisture content at the farmer field level. However, also when applying models to smaller field scale catchment, no satisfactory results were obtained in the prediction of measured surface and sub-surface runoff. There can be different reasons, but the main ones are related to the spatial-temporal variability of soil hydraulic properties used as input to process based models encountered in Norwegian soils. This problem however can partly be overcome when using spatially valid data for the soil moisture content, instead of calibrating the model using individual soil profile data. In the IRIDA project, the Hydrus-1D model was used applying an inverse modelling technique, resulting in more reliable estimates for the soil moisture content. Through the project increased knowledge in the use of models has been obtained in combination with satellite imagery. Benefits will be related to potentially more precise estimates of flash flood warning, runoff and nutrient and soil loss and soil moisture content facilitating farming operations, aspects definitely needed in a future bioeconomy and conditions of climate change. But additional research related to model choice, satellite imagery interpretation and the inclusion/use of temporal and spatial variable input parameters is needed. In this respect, a project proposal dealing with the variability and modelling using the HYDRUS model, has been submitted to Landbruksdirektoratet (LDIR). Also, cooperation is planned with the Romanian partner through funding by the EEA/Norway grants. and with the University of Catania, Italy through the Erasmus program KA103 - Mobility for Traineeship.
The project has resulted in knowledge about the various uses of drones and is as of now used in new projects related to the mapping of gully erosion, planning of retention basins and quantifying streambank erosion. Drone knowledge is also used for educational purposes, among others as part of project cooperation with the Kjelle Videregående skole. A bachelor student from the Netherlands was taught in the operation of drones and using drone imagery to identify rill erosion on a 3D elevation model. Knowledge in using satellite imagery has been obtained mainly through the IRIDA project now facilitating accessing, processing and analysing satellite imagery timeseries. This new knowledge has among others resulted in the development of a project proposal with Kosovo, related to mapping pollution along riverbanks. Imagery is now also used to locate potential water-prone areas in agricultural landscapes and in the identification/mapping of farmer field boundaries in agricultural dominated catchments.
Gjennom IRIDA prosjektet er det oppnådd;
Økt kunnskap om bruk av droner og satellitt til registrering av vekstutvikling og jordas vanninnhold
Økt kunnskap for å gi informasjon om riktig tidspunkt for jordarbeiding for unngå jordpakking
Økt kunnskap i bruk modeller til simulering av jordas vannbalanse, avrenning og tap av næringsstoffer fra landbruksdominerte nedbørsfelter
Økt kunnskap om jordas vanninnhold gjennom bruk av data fra satellitt i kombinasjon med modeller bidrar til bedre flomvarsling
Innsamling av informasjon om jordas vanninnhold kombinert med modeller kan anvendes i DSS (Decision Support System) for Norske forhold.
Bedre kunnskap i anvendelse av modeller bidrar i bedre simulering av effekter av klimaendringer på avrenning og tap av næringsstoffer og dermed valg av riktig
Økt kunnskap om bruk av droner og satellitt anvendes i diverse nye prosjekter, både nasjonalt og internasjonalt
Økt kunnskap om bruk av droner blir brukt i undervisning
Efficient agriculture water use is of crucial importance for water resources management. Consequently, accurately determining evapotranspiration (ET) is the first step for improving irrigation efficiency and productivity and for quantifying the ecosystem water balance. Several approaches for determining ET have been proposed in literature, but the relation between high and low spatial resolution methods still remains unresolved. This proposal will create a mixed model where isolated actual ET and soil moisture measurements can be correlated with actual ET results obtained by means of low-resolution methods. The combination of on-the-ground high-resolution ET methods with the analysis of thermal and hyperspectral imagery provided by unmanned aerial vehicle (UAV/RPAS/UAS) (at plot scale), manned vehicles and satellites (at catchment scale) should ease the mixing performance and solve the upscaling. The proposal will integrate the methodologies and routines into a decision support system (DSS) that will serve to manage large amount of inputs (Big Data Analysis) and provide simple irrigation recommendation to the end-users. At plot level IRIDA will set, through the analysis of high-resolution thermal and hyperspectral imagery provided by UAVs, the range of variability to detect water stressed zones. This information will be used to decide the exact location for installing on-ground sensors to increase the spatial representativeness of the ET. At a catchment scale under conditions of varying land use as in northern Europe, the evaluation of satellite remote sensing will allow increasing the accuracy of the ecosystem water balance determination, improving flood predictions and the water footprint assessment. The obtained results will be disseminated at a scientific level and a market exploitation study will be carried out by the public/private partnership from 4 different countries representing the great diversity of agro-systems and their water management in Europe