from large remote sensing time series can be exploited in a Big Data framework
to improve forecasting models of eco-hydrological variables at regional scales,
understand vulnerability of ecosystems
climatic extremes (droughts, temperature), quantify the effects of land use changes on
the hydrological cycle and forecast agricultural products
the project, machine learning algorithms (e.g. tensor decomposition, random
forest) can be used with MODIS, Sentinel, OCO-2 (plant
or GRACE satellite datasets together with climatic products (ECMWF).
be considered depending on
gap filling of time series due to cloud cover, estimate risk of fires, forecasting of prices
wine and olive oil from high value regions,
forecasting of crop yields or water
is part of the
European project. Collaboration with companies is possible (DHI-Gras).
I samarbejde medDHI Gras
programming in Python, Matlab