Mapping UK Habitats

A soil carbon and land use database for the United Kingdom.
[Bradley et al 2005]

soils-uk

www.landis.org.uk

The above paper describes the compilation of a database to estimate soil carbon stocks and carbon dioxide emissions from UK soils by interpolating the analysis of 11,000 soil horizons and site data with legacy soil maps at 1:250,000 scale  in order to

derive high-resolution spatial data on soils and land-use data for use by a dynamic simulation model of carbon fluxes from soils resulting from land-use changes”.

 

Since the creation of the database the European Space Agency’s (ESA) Copernicus missions have measured, mapped and observed the Earth’s surface at 10, 20 and 60m resolutions. A vast library of images is now freely available, one that could be used to improve and update the habitat component of the soil carbon and land use database above. Furthermore having used an extended spectrum the data not only provides imagery in the visible spectrum but also in the infrared and microwave to provide data on the thermal, hydrological and gaseous properties of the Earth’s surface. It’s a library that could significantly enhance and updated the soil carbon and land use database to the extent that it extends its use beyond CO2 simulation models.

The Crowd

AfSISIn Africa the African Soil Information Services (AfSIS) has been taking advantage of this imagery to develop land use maps. It’s a first for Africa which doesn’t have the underlying soil maps to build upon and similarly has over 30 million km2. Working at a resolution of 250m2 AfSIS are using a simple yes/no question analysis to utilize the power of the crowd to map Africa’s land cover. Questions, that as with Bradley, set out to identified the basic land uses in Africa: Cultivated, Grassland, Woodland and Built. Categories that correspond to Bradley’s categories of Cultivated-Arable, Woodland, and Semi-natural.

It’s worth noting that as the UK has only 240,000km2, (under 2 million images at a resolution 250m2), it would take two thousand volunteers less than a month to completely revise and improve the resolution,by a factor of eight, and accuracy of the land cover aspect of the soil carbon and land use database. As the Copernicus missions are ongoing so could be this revision process so that the actual land cover is always accurately reflected in the database.

Farm/Field derived Data

Approximately 70% of the UK’s land is agricultural with half under cultivation.

data sources for mapping uk habitatsIn many cases these cultivated lands are utilized in precision farmer operations where soil properties have been measured and mapped at a resolution of 1ha or less. With as much as 1/3rd of the UK’s soils properties mapped at field scale, and the legacy soil maps used by Bradley at a resolution of 1km, these field measurements represent a great opportunity to enhance the soils data in the soil carbon and land use database.

The British Geological Society (BGS) have similarly produced a mysoil app which allows the general public to asses their soil and add data to the BGS database. Unfortunately there is no GNU/Linux version of this app and my request for more information has not yet been replied to so I can only assume that this app could also help to improve the soil carbon and land use database.

Citizen Science

The advent of the internet has bought about the opportunity for mass data exercises that were previously impossible for a small team of researchers to analysis extensive. Thus studies relied on supporting datasets by using statistical techniques designed to squeeze out the biggest truth from the least amount of effort. With the internet there is no need to use these support methods to the same extent and instead the statistics can be applied to qualify the integrity of the data and illuminate its implications. (link: The BGS Citizen Science home page)

 

The Green Data Revolution

sentinel-2aFrom Satellites to Termites…

Flying high above the Earth’s poles are the two satellites of the European Space Agency’s Sentinel-2 program. Part of the Copernicus initiative  they are mapping Africa at 10m, 20m and 60m resolutions so as to  provide data to facilitate the creation of accurate maps for environmental management and to study the effects of climate change.

Root zone, Soil Moisture http://www.esa.int/

For more than 30 years the ESA’s Living Planet Programme (Explorer missions, Earth Watch and Copernicus Sentinel missions) has remotely sensed and created a data archive on the Earth’s  climates, biomes and the processes that operate within them.

Mapping Africa’s Habitats

AfSISThese satellite archives are now being utilized by  The African Soil Information Service (AfSIS)  to describe Africa’s soil and landscape resources. It’s a project that will contribute to the development of sustainable agricultural systems that can feed the populace and co-exist with Africa’s wildlife.

Mapping 30 million km2 of the Earth’s terrestrial surface at a resolution of 250m2 is though a mammoth task that can only succeed with the effort of the crowd. So if you care about the elephants and lions, lend AfSIS your eyes, map a few km of Africa and help to make the World sustainable.  Join the crowd and map Africa’s habitats.

Digitizing the Herbarium 

Sustainable agricultural strategies need to be comprehensive and include details of the plants and animals that inhabit the land. This is particularly so for agricultural where poor crop choices and cultivation techniques can quickly and irreversibly rob the land of it’s soil, and water.  A situation that forces farmers to fell virgin forest and plough up natural grasslands in a repetition of the exercise. Many of the World’s deserts, including the largest, the Sahara, have been significantly expanded by this process of agricultural degradation.

Solanaceae-sourceProjects such as the International Plant Names Index, the World Flora online and the Solanaceae source are in the process of constructing comprehensive botanical databases for the 350,000 plant species that inhabit this planet.

Accessing Crop Data

With less than 100 plant species responsible for over 80% of the World’s raw material and food production, the digitization and access to bibliographic and research data on crops is essential if we are to achieve any semblance of agricultural sustainability. To this ends The ODI (Open Data Institute)  and GODAN (Global Open Data for Agriculture and Nutrition) are both exploring what the challenges are and what the global priorities should be.

A Digital Zoo

nhmThe Natural History Museum are digitizing over 80 million specimens from their avian, entomological ,and botanical collections. Many of these collections include additional physiological and habitat data such as the Hostplants and Caterpillars  and the Host-Parasite databases.

Animal Husbandry 

Just as we cultivate a fraction of the Earth’s plants so we have domesticated fewer than 30 of its animals. The agriculturally significant have though been diversified into many and varied breeds each with it’s own characteristics and habitat requirements. I know of no plans to digitize and make details on the requirements, management, or impact of different livestock breeds and husbandry methods on the environment despite this information existing in abundance.

Crop Pests

plantwiseCABI (Centre for Agriculture and Biosciences International) is the secretariat for GODAN and manages the plantwise project. Through the plantwise project CABI run clinics on pest and disease management and have similarly developed a searchable knowledge bank for pest identification.

Smart Data

Making use of this data and turning into something practical and useful was the remit of smartopendatathe smartopendata project. However both the web site and the final report are littered with unexplained acronyms [wtf?] and composed in a language that could well have been uttered by a Lewis Carroll character: “Hereafter, using this central core, the pilots extended the SmartO penData vocabulary to take into a account their own singularities. ”

That said, a Singularity, as in a technological/ digital singularity, is perhaps the correct adjective and conceptual framework to ‘think’ in when talking about smartdata for Agriculture and the Environment.

Let the Data Make the Decisions?

The Singularity in this sense is not an AI (artificial intelligence) but a singular objective [sustainability], within a single integrated application environment. An application environment that extends beyond basic crop and animal husbandry advice into a fully integrated service that can calculate and allocate water resources, greenhouse gas emissions or the potential for erosion. Services that can be further integrated with economic and socio-political objectives so that land and crop choices are efficiently made to meet market needs without over and under production. The same mechanism could be used to reward/compensate for mitigation strategies in the name of climate change and biodiversity objectives.

A digital singularity doesn’t  have a persona, a corporate or a national identity. With no concept of self it seeks the optimum solution based on the data supplied. The more comprehensive and far reaching that data the better the decisions that can be made using the singularity. A singularities decisions are however only as good as the data supplied; inaccurate, incomplete or misleading data, just as it does in real life, can lead to a catastrophe.

The Need for Education

Supplied with comprehensive and accurate data, a digital singularity can encourage sustainable practices, but for them to be adopted correctly a farmer needs to understand the reasoning behind the action. If global targets for climate change and habitat preservation are to be met then farmers need to be informed of the significance, understand the logic and gain tangible benefit from implementing a given strategy. Strategies that similarly require feedback (data input) from the farmers and that can only be achieved with the informed cooperation of those farmers, many if not most of which are illiterate.

 

Next Part II