Because of insufficient accessible arable land, intensive farming has been linked to excessive accumulation of phosphorous, heavy metals, and other soil contaminants, as well as to significant groundwater pollution with nitrate. Deterioration of soil water quality is especially worrying at the bioclimatic Mediterranean area, especially under the current context of climate change. Hence, it is necessary to develop a common body of knowledge, shared at the local and regional levels of the countries involved and affected, so as to allow an effective monitoring of cropping systems, fertilization and water demands, and impacts of climate change, with a focus on the sustainability and the protection of the physical environment.
In this presentation, we describe AgriBigCAT, an online software platform that combines geophysical information from various diverse sources, together with big data analysis, in order to estimate the impact of the agricultural sector on the environment, considering land, water, biodiversity and natural areas requiring protection, such as forests and wetlands. Based on the P-Sphere project, this platform intends to promote more sustainable agriculture, by designing and developing an information and knowledge-based platform, using a big data approach for managing and analyzing a wide range of geospatial and mainstream information, which can be accessible by standard communication technologies such as the internet/web and mobile apps. this platform can also assist both the farmers' decision-taking processes and the administration planning and policy making, with the ultimate objective of meeting the challenge of increasing food production at a lower environmental impact.
CSR_Module5_Green Earth Initiative, Tree Planting Day
Estimating the Impact of Agriculture on the Environment of Catalunya by means of Big Data Analysis
1. 1
P-Sphere Project:
Estimating the Impact of Agriculture on
the Environment of Catalunya
by means of Big Data Analysis
Andreas Kamilaris
17th February, 2017
2. Problem
2
Intensive farming linked to excessive accumulation
of nutrients (N and P) and contaminants (metals,
POPs) in the soil.
Significant groundwater pollution with nitrate.
Emission of acidifying and greenhouse effect gases.
Deterioration of air/soil/water quality are worrying at
the bioclimatic Mediterranean area, especially under
the current context of climate change.
3. Motivation
3
Need for a common body of knowledge.
Shared at local and regional levels of countries
involved and affected.
Effective monitoring of cropping and animal
production systems, fertilization and water demands.
Estimations of impacts, including climate change.
Focus on sustainability and protection of the physical
environment.
5. Internet of Things
5
A network of objects, where all the physical “things” are
uniquely and globally addressable, manageable and
identifiable by computers, in the same way like by humans.
7. 7
Big Data
Volume: The size of data collected for analysis
Velocity: The time window in which data is useful,
accurate and relevant.
Variety: Multi-source (e.g. images, videos, sensing data),
multi-temporal (e.g. collected on different dates), and
multi-resolution (e.g. different spatial resolution images).
Veracity: The quality, reliability and potential of the data,
as well as their accuracy, reliability and confidence.
Valorization: The ability of big data to propagate
knowledge, appreciation and innovation.
8. 8
Big Data: Sources for Agri
• Cameras
• GPS sensors
• Physical sensors
• Weather stations
• Remote sensing from drones and other UAV
• Remote sensing from airplanes and satellites
• Web data from online web services
• Feeds from social media
• Crowdsourcing-based techniques from mobile phones
• Static historical information: databases and statistics
• Humans as sensors
9. 9
Big Data: Analysis Techniques
• Image processing
• Machine learning
• Cloud-based Platforms for large-scale information storing,
analysis and computation
• Geographical information systems (GIS)
• Big databases
• Message-oriented middleware
• Modeling and simulation
• Statistical tools
• Time-series analysis
10. Project P-Sphere: Research Questions
10
How can we accurately measure the
environmental impact of agriculture in Catalunya
using big data analysis?
Which solutions can we find to avoid the negative
effects of animal manure on the environment?
11. Project P-Sphere: Methodology
11
1. Collect datasets from Internet of Things sensors
used in agriculture and weather monitoring.
2. Develop a Big Database for storing this information
for easy retrieval and analysis.
3. Use the datasets as layers into a geospatial analysis
tool/application.
4. Apply Big Data Analysis to estimate environmental
impact and find viable solutions.
5. Enhance analysis with real-time info from Web of
Things sensors (e.g. weather, hazards, alerts).
12. Project P-Sphere: Data Sources
12
• Farmers & Animal types/numbers
• Climatic conditions (temperature, humidity, evapotranspiration)
• Infrastructures (transportation network, pipelines system)
• Areas of natural interest, areas that require protection
• Forests
• Agricultural parcels
• Air quality
• Soil characteristics
• Manure management units
• Statistics of the population
• Biodiversity (animals, birds, micro-organisms)
• Water (lakes, rivers, precipitation)
13. Project P-Sphere: Specific Goals
13
• A complete geo-information
inventory/model that could be
used by agriculture scientists
• Circular economy – waste
management
• Quantify impact – focus on
manure management
• Propose solutions based on
ICT technologies
• Examine “what if” scenarios
to mitigate/avoid impact.
14. Project P-Sphere: Actual Status
14
• Collection of data sources – use as layers
• Geospatial application
Data
layers
Tools for
spatial
analysis
GIS
visualization
15. Project P-Sphere: Actual Status
15
Cultivations per municipality Stations of meteorology and manure management
Forests and annual precipitationTransportation and pipelines network
16. Project P-Sphere: Actual Status
16
• Calculation of animal manure produced annually in Catalunya.
• Estimation of gases produced:
• Carbon dioxide, Methane, Nitrous oxide
• Ammonia, Odor
• IPPCC (TIER1) Vs. Relevant Literature (TIER2)
17. Project P-Sphere: Actual Status
17
All together as a Web of Things application
Emissions
calculator
Data
layers
GIS visualization in
the Web browser
Farms
involved in
the results
Weather
conditions and
forecasting
24. 24
Project P-Sphere: Big Data Aspects
Volume: Datasets, layers, estimations/calculations, spatial
analysis.
Velocity: Weather information, precipitation patterns.
Variety: Different sources of information involved, i.e.
historical data, satellite data, real-time web feeds, Internet
of Things sensor data.
Veracity: Trusted source/origin, i.e. Ministry of Agriculture,
AccuWeather predictions, international satellites, research
project outcomes.
Valorization: Analysis, simulation, modeling.
25. 25
Project P-Sphere: “What if” scenarios
Volume + Valorization:
What if we implement Best Available Techniques (BAT) for
manure treatment in certain areas? What are the benefits/costs?
What if we construct centralized manure processing plants?
Where? How many? To which extent they reduce transportation
costs and environmental burden? When is the expected ROI?
What if we give incentives to the farmers to produce/use organic
fertilizers? Could these incentives fluctuate like stocks in a stock
exchange system, as replacement for non-renewable sources?
What if we monitor through satellites the manure management
handling at each farm? Is it feasible? How about nitrogen? How
much effort is needed in terms of data storage and computation?
26. 26
Project P-Sphere: Big Data Aspects
Example: Create density zones for agriculture
What about more complex questions in specific time window?
Locally: ~20 minutes
Cloud: ~4 minutes
29. 29
Project P-Sphere: Next Steps
More datasets and layers
Better estimations/calculations
Geospatial analysis
Examine various scenarios.
Adaptation to the particular needs of the Ministry of
Agriculture.
Use of Big Data technologies to improve performance and
increase scalability.
30. 30
Many thanks for your attention!
Andreas Kamilaris
andreas.kamilaris@irta.cat