The overarching goal of HIGHLANDER is to support a smarter (environmental and economically sustainable) management of lands in their mosaic of natural resources, land uses, sectors, human activities and assets, while also reducing risks and taking opportunities posed by climate change and its variability.

HIGHLANDER has three main interrelated objectives:

  1. to design and implement a framework of multi-thematic, last generation, continuously updated, highly detailed and harmonized data, indicators and tools, from remote and in-situ monitoring, analytical tools, numerical models up to machine learning algorithms.
  2. to fully exploit the HPC capabilities to generate, post-process, host, distribute and make accessible to and exploitable by multiple users both existing and newly generated data, indicators and tools into HPC-based services enabling the mainstreaming of the information itself into decisions, strategies and plans at different interacting spatiotemporal scales and sectoral levels.
  3. to ensure the long-term functionality of the created services thanks to the involvement of real users during the project. Huge amount of input data gathered and used in HIGHLANDER will come from geospatial and non-geospatial datasets. From these, new data, indicators and tools will be generated, harmonized and made available to be then efficiently exploited by multiple end-users, e.g. farmers and their associations, forest/ecosystem managers, water providers, entrepreneurs, service providers, policy makers, practitioners, educators, researchers, government.

The foreseen services are:

  • a National Open Data Portal on climate and climate-related hazards.
  • a set of illustrative “downstream applications”, aimed mainly at increasing the users’ awareness about the great potential of the information provided as well as their capacity in further elaborating and using data;
  • implementation of a chain of operations that directly generate the user-required information through a usercustomized package.

The consortium of partners will define open data policies for the gathered or newly-generated data in order to define a national regulation and their re-use.

Moreover, this Portal will be integrated into the European Data Portal to achieve a highest level of visibility through a pan-European synergy.

Resource and data analysis

Cineca is committed to a two-way participation by providing both a high-end cloud service for data analysis and practical support for specialized research problems through machine learning techniques. A virtual server consisting of expandable storage (currently ~ 200 GB) and multicore CPU Core Core Processor (Broadwell, IBRS) was dedicated within the recently updated cloud framework available at Cineca. The server offers a fully-fledged centralized platform for housing data and creating complete machine learning pipelines ranging from data preprocessing to model evaluation and optimization, based on the needs of end users. The service is equipped with all the necessary software tools, mainly based on Python and R, for the prototyping and testing of machine learning protocols. In addition, Cineca also joins other partners, research professionals and experts, to address domain specific research problems of high computational complexity.

Climate effects on biodiversity
The examination of environmental pressures on biodiversity, especially in extreme conditions, is an important problem that can provide useful insights on the direct influence of external parameters on the delicate balance between biological ecosystems. The project aims to study, on the basis of an integrated multidisciplinary approach, the distributions of endolithic communities and the biodiversity of the mushroom and bacteria population in different environmental conditions such as altitude and exposure to the sun. By integrating biological, geological and microclimatic data, the habitability limits on various regions of interest will be decoded.

Satellite image analysis
Remote sensing is a well-established technique for long-range detection and perception of target positions. The images collected with these techniques are composed of a very narrow continuous spectral band with hundreds of spectral bands, which cover bands of visible, near infrared, medium and thermal infrared light regimes, depending on the context. The information hidden in the spectral signatures contained in these images is extremely useful for discovering and monitoring, over time, environmental changes in the target surfaces. The project aims to create a prototype machine learning pipeline for processing satellite images. To this end, the images containing radar and spectral data are analyzed through automatic learning protocols to obtain insights on the variations of the green coverage in different target locations of interest.

Nine Use Cases are currently part of the project

  • Land suitability for vegetation (forests, crops) - Starting from the current distribution by latitude and altitude of forest and harvested species, we will try to build a forecast model of their possible migration as a result of climatic evolution.
  •  Human wellbeing in rural and urban areas - Starting from the current distribution of populations, we will try to study their quality of life, with particular attention to the differences between urban and rural areas.
  • Water cycle and sustainability of competing uses (hydropower, domestic, agriculture, ecological) - Starting from a study conducted in the catchment area of ​​Ofanto, in Puglia, the idea is to extend the methodology to other basins, creating forecast models on use of water resources in the energy, irrigation and domestic sectors, guaranteeing at the same time the vital flow for the basin itself.
  •  Soil erosion - The general objective is to obtain an empirical model for soil erosion forecasts, both following natural events (forest evolution, landslides, landslides) and agricultural activity. The aim is therefore to recommend targeted exploitation, also in terms of soil tillage and planting methods.
  • Forest fires prediction - By means of satellite images and remote sensing activities, the project aims to monitor fire risks.
  •  Natural parks environmental management - Using satellite images and remote sensing activities, the initial idea of ​​the project is to relate the animal movements and habits of the herds grazing in the Parco del Tentino with the vegetation present while simultaneously monitoring the risks of fire.
  • Crop water requirements forecast - Except for special cases, irrigation takes place mainly in the summer. Due to climatic conditions other than traditional ones, the need to irrigate begins in spring and during the sowing phase. The challenge in HIGHLANDER is to test the insertion of medium-term (quarterly) climatic data in the model currently in use and test the reliability of irrigated forecasts.
  • IoT for Animal wellbeing - High temperatures, as well as strong temperature fluctuations cause high stress to animals, both in the pasture and in the stables. The collection of data on animals and breeding environments and satellite observations integrated with weather forecasts aim to improve animal welfare.
  • Animal Welfare and land suitability for farming - The exploitation of new bioclimatic indices developed from data and medium-term climate indices will be used to develop models to support farmers in the management of extreme weather events and predict the stress of plantations later to events such as hail or drought to limit damage, optimize the use of resources (water and energy), prevent destructive fires and monitor landslides and landslides of cultivated land.
data chiusura progetto
Partner
  • CINECA (Coordinator)
  • Agenzia Regionale per la Prevenzione, l'Ambiente e l'Energia dell'Emilia-Romagna (Arpae Emilia-Romagna)
  • Agenzia Regionale per la Protezione Ambientale del Piemonte
  • ART-ER
  • CONFEDERAZIONE ITALIANA AGRICOLTORI PROVINCIA DI TORINO C.I.A
  • Fondazione Centro Euro-Mediterraneo sui Cambiamenti Climatici (CMCC)
  • DEDAGROUP Public Services S.r.l.
  • ECMWF
  • Fondazione Edmund Mach
  • Department for innovation in biological, agro-food and forest systems (DIBAF) Universita' della Tuscia

 

Finanziamenti e tempistiche

 

Co-financed by the Connecting European Facilty Programme of the European Union Grant agreement n° INEA/CEF/ICT/A2018/1815462

 

Ambito
Unione Europea