Perspectives from a Custodian Agency for Agriculture, Forestry and Fisheries
Food and agriculture lie at the heart of the 2030 Agenda, with closely related development outcomes that range from ending poverty and hunger to maintaining and protecting the natural resource base, and responding to climate vulnerability and change. As a result, FAO was chosen as the Custodian Agency of 21 SDG indicators, with responsibilities for the methodological development, the provision of technical assistance and the collection and dissemination of data for monitoring progress towards a number of targets under Goal 2 ‘Zero hunger’, Goal 5 ‘Gender equality’, Goal 6 ‘Clean water and sanitation’, Goal 12 ‘Responsible consumption’, Goal 14 ‘Life below water’ and Goal 15 ‘Life on land.’ Data collected from countries and new sources will allow monitoring annual progress at a sub-regional, regional and global level and will provide the evidence base for the planned follow-up and review processes in the context of the SDG High Level Political Forum.
5.1 Tracking progress towards sustainable development
Earth observations (EO) can provide a significant contribution to the measurement of many of the SDG indicators under FAO custodianship. In particular, remote-sensing images and georeferenced data can support the design and development of more efficient and accurate sampling frames in the preparation of integrated agricultural surveys used for monitoring SDG indicators. Secondly, area changes in natural vegetation assessed from satellite imagery directly inform the measures of specific indicators, such as the Green Mountain Index. Thirdly, EO stratified by land cover information are useful auxiliary variables to enhance data coherence and accuracy. For instance, satellite imagery may complement ground observations for computing critical sub-components of more complex indicators, such as the measurement of the area under sustainable and productive agriculture. Finally, EO are key data for spatial disaggregation, including for the distinction of rural and urban areas required for the computation of several SDG indicators.
Accordingly, FAO’s support to countries, in the context of the SDG indicators, increasingly includes capacity development activities based on geospatial tools. This article provides examples of the applications mentioned above, as implemented by FAO, including with support of the Global Strategy to Improve Agriculture and Rural Statistics, demonstrating the specific role EO can play in helping countries meet the monitoring challenges ahead.
5.2 Farm-based surveys and the use of geospatial information
Progress towards achieving Target 2.3 and Target 2.4 of Goal 2 is measured by three global indicators that are meant to be informed by agricultural surveys whose statistical unit is the farm. Target 2.3, in particular, focuses on the economic performance of small-scale food producers, measured by their income and productivity:
- indicator 2.3.1: volume of production per labour unit by classes of farming/pastoral/ forestry enterprise size;
- indicator 2.3.2: average income of small-scale food producers, by sex and indigenous status.
Target 2.4 focuses on the sustainable increase of agricultural productivity:
- indicator 2.4.1: proportion of agricultural area under productive and sustainable agriculture, which entails maintaining agriculture’s ecosystems function, by improving land and soil quality and strengthening its capacity for adaptation to climate change, including improved resilience to extreme events and disasters.
Monitoring this target involves measuring the economic, social and environmental dimensions of agricultural sustainability with appropriate sub-indicators.
The official global indicators selected to measure progress against targets 2.3 and 2.4 require a common data collection framework, able to gather timely and relevant environmental, economic and social information at the farm level, with the possibility of capturing disparities between small- and large-scale food producers.
In order to better meet these requirements and more generally the need to improve the quality, consistency and timeliness of national and sub-national agricultural data, FAO has recently proposed a new approach to agricultural surveys, the Agricultural Integrated Surveys (AGRIS), which aims to gather information on both the core activities and the key characteristics of the farm, in particular those that will be needed for monitoring of SDG indicators 2.3.1, 2.3.2 and 2.4.1.
EO data, in particular satellite imagery and ortho-rectified aerial photographs, together with geo-referenced information are essential tools in designing a consistent, efficient and well-integrated sampling frame for AGRIS in order to enable sampling and reporting with equal efficiency at farm, household and landscape scales, with the ability to link information across multiple thematic domains. Stratification of satellite imagery by relevant land cover strata improves the sampling efficiency of agricultural surveys, with respect to both types of area and list sampling frames, which are typically used jointly for agricultural purposes. The use of satellite imagery also supports and increases the efficiency of ground work, facilitating ex-post data corrections needed to improve quality control of the survey estimates.
Figure 2: FAO’s Open Foris is a set of free and open-source software tools that facilitates flexible and efficient data collection, analysis and reporting.
Figure 1: The 21 SDGs indicators under FAO custodianship
5.3 EO data for direct monitoring of SDG indicators
In response to the SDG monitoring needs, FAO has stepped up its own efforts to exploit cutting-edge technologies designed to access and analyse information on land and natural resources from remote-sensing sources. For instance, FAO has developed the Open Foris suite in partnership with Google. Open Foris is a set of open-source software tools, including Collect Earth in particular, that are instrumental to the data measurement of several indicators especially relevant to Goal 15 ‘Life on land’.
Within goal 15, SDG indicator 15.4.2 focuses on a wide range of universally important services provided by mountain ecosystems, as a basis for sustainable mountain development. The indicator’s methodology focuses on measuring changes in the area of green vegetation in mountain areas (forest, shrubs and pasture land, and cropland) as a proxy for changes in ecosystem function of mountain environments. FAO supports monitoring of indicator 15.4.2 “Mountain Green Cover Index” through a customized application of Collect Earth. Collect Earth was applied to extract index values disaggregated by country, elevation class and IPCC land use categories and to compile them in a 2017 baseline. Changes in mountain vegetation over time will be assessed against this baseline.
5.4 EO as complementary variables for national assessments
The indicator 15.1.1 “Forest area as a proportion of total land area” measures the status of conservation or restoration of forests in a country, indirectly contributing to measuring to what extent they are sustainably managed. Changes in forest area may reflect changes in demand for other land uses due to economic activity and pressures. To this end, this indicator provides crucial information for policies in support of sustainable forest and landscape planning. To monitor forest cover and changes, EO is increasingly complementing the data that FAO has historically collected through the Forest Resources Assessments (FRA). Offering better access to satellite imagery and to tools for image processing and data interpretation, new FAO applications such as Collect Earth and the System for Earth Observation Data Access, Processing and Analysis for Land Monitoring (SEPAL) are contributing to improved forest monitoring, complementing more traditional collection of national data through questionnaires.
In the context of indicator 2.4.1, high-resolution imagery contributes to assessments and mapping of soil organic carbon (SOC) at farm scale, as part of regression models and as a source of land use stratification. SOC is a critical aspect of soil health, which is in turn one of the components used to assess the environmental dimension of agricultural sustainability.
With regard to indicator 14.6.1, monitoring systems housed on fishing vessels and based on satellite data are being proposed for tracking illegal fishing activities in real-time and could significantly contribute to FAO’s efforts to combat illegal, unreported and unregulated fishing (IUU).
Coherent frameworks of data collection, monitoring and reporting can stimulate synergies among UN agencies and with national statistical authorities. The use of EO has been instrumental in building these synergies for indicator 15.3.1 that monitors the status and trends in land degradation. The UN Convention to Combat Desertification (UNCCD) is the Custodian Agency for this indicator, but FAO supports the monitoring of one component of this indicator by contributing its expertise on land-related statistics. FAO has traditionally led the development of international standards for land-use and land-cover classifications, such as those adopted in the 2020 World Programme for the Census of Agriculture (WCA 2020) and the System of Environmental-Economic Accounting (SEEA) Central Framework. In addition, FAO coordinates a long-standing reporting process on land-use information from member countries, which may use remote-sensing land cover mapping for the validation of national data. In order to support this process, FAO has recently developed reference statistics based on global land cover maps, disseminated via FAOSTAT (www.fao.org/faostat/en/#data/LC).
Finally, FAO is now partnering with the European Commission, OECD, the World Bank, the Global Strategy to Improve Agricultural and Rural Statistics, and other UN organisations to develop an agreed international definition of urban and rural areas for consistent reporting of SDG indicators and beyond. The methodology classifies Local Administrative Units on the basis of a combination of criteria of geographical contiguity, minimum population thresholds and economic activity of the resident population applied to 1 square km population grid cells. Medium to very high-resolution imagery are the source of existing global land cover maps and human population distribution layers that underlay the proposed methodology to distinguish rural and urban areas globally. More on these aspects may be found in the UN-Habitat article here in Part II.
Figure 3: Built on Google desktop and cloud computing technologies, Collect Earth facilitates access to an unparalleled amount of freely available archives of satellite imagery, including very high resolution and frequency imagery. Collect Earth streamlines the use of probability sampling offering a robust and fully customizable framework for data collection. It allows the capture of new information on agriculture and natural resources for monitoring targets across the 2030 Agenda, from crop monitoring to land and forest cover, from pest/locust control to water management, from plant health to losses due to natural disasters.
5.5 Country and international data for global reporting: challenges and opportunities
Reaching the goals and targets of the 2030 Agenda for Sustainable Development requires the establishment of global monitoring and reporting processes. These processes should be based as much as possible on national data in order to ensure country ownership. In some instances, however, international agencies may use non-official data to construct international data series in fields that are not covered by existing official sources, or where a single source (e.g., EO satellite imagery) may provide more consistent and lower-cost data to measure a global or trans-national phenomenon than the results of amalgamation of multiple individual country datasets.
Non-official sources might sometimes also be used by international agencies to estimate country-specific values of SDG indicators when national official data do not exist, are incomplete or do not comply with international standards; or to impute missing values within a national official time series or to extrapolate official time series. In this respect, land, water and agri-environmental statistics derived from satellite imagery support the construction of a consistent data framework across sub-national, national and global scales.
Figure 4: FAO and Global Strategy guidelines on applying remote sensing information to improve crop statistics.
As a result of this work, discrepancies may arise between international and national estimates of similar SDG indicators. This may be a cause of concern for some national authorities given the reputational risk for countries to have their data contradicted by those published by international organizations.
International organizations can address these concerns by strengthening the statistical capacity of countries in areas where data are not available or not compliant with international standards, with the goal of enabling them to produce their own data in the long run. FAO’s activities towards improved national statistics of its member countries are an integral part of this effort. EO and Big Data complement local knowledge and expertise and can boost the efficiency, quality, transparency, credibility and above all the timeliness and efficacy of data collection and the validation of existing global products.
Applications based on remote-sensing data play a major role for building statistical capacity in countries and for promoting knowledge sharing at the regional level. FAO is offering training on the use of FAO software tools such as Open Foris and Collect Earth to national experts who will be able to conduct – in a few hours – mapping and classification exercises that used to take weeks or months.
Article Contributors
Pietro Gennari, Francesco N Tubiello and Giulia Conchedda (Food and Agriculture Organization of the United Nations)
Further Information
FAO, Office of the Chief Statistician: chief-statistician@fao.org
