Scientific Experiments

COLLABORATE ON OUR

Scientific Experiments

The Virtual Lab grants access to the experiments performed in all scientific domains of the PrimeWater project: remote sensing, process-based modeling, and data-driven modeling. The Virtual Lab describes in detail the data, equipment, and experimental procedures followed in each experiment to enhance transparency and secure the reproducibility of results. Nonetheless, the Virtual Lab is not a mere repository of experiments; it welcomes and embraces collaborative research efforts across organizational boundaries. The Virtual Lab facilitates research groups to exchange knowledge in a collaborative framework that will ultimately advance remote sensing and hydro-ecological modeling.

Remote Sensing

A1. Imaging spectrometry

Explore current imaging spectrometry data (e.g. PRISMA, DESIS) in order to describe in terms of pigments, size classes, and functional traits of primary producers.

A2. Multiscale Optical Data

Explore concurrent measurements from in situ, drone, airborne and satellite sensors

B1. Assimilation of Earth Observation products and in-situ data in seasonal hydrological forecasting services

Best practices for seasonal forecasting

B2. Assimilation of Earth Observation products in water quality modelling

Can we combine EOs and 4dVAR data assimilation technique, to improve the predictive skill of hydro-ecological modelling in reservoirs?

B3. Limits of one-dimensional process models to forecast water quality

Quantifying the relevance of the major sources of uncertainty and establishing the forecasting horizon of one-dimensional process models

B4. Developing an error-correcting complementary modeling approach for surface water quality

Can we train data-driven algorithms to accurately detect systematic errors produced by process-based models?

C1. Assess the relevance of hydrometeorological variables in phytoplankton dynamics of surface waters using data-driven modeling

Can we use data-driven algorithms to gain insight in the drivers of phytoplankton dynamics in lake and reservoir ecosystems? Can we reduce the predictor set? What do we gain in terms of accuracy and uncertainty from this reduction?

C2. Develop and benchmark data-driven algorithms to forecast the short-term dynamics of phytoplankton in surface water reservoirs

How much more accurate can data-driven algorithms be in predicting water quality compared to naïve predicting alternatives? How do they compare to each other in terms of accuracy and uncertainty? What are the limits of predictability?

C3. Quantile-based modeling of phytoplankton dynamics using Random Forests

Can we improve the predictability of such events that occur at the limit of probability? If we can, how reliable is the prediction for a new instance?