Remote sensing in development cooperation
Framework conditions and recommendations for action
Successful use of remote sensing in projects
Project owners face the challenge of incorporating remote sensing into the project planning and ensuring it is used effectively – potential uses are often unknown and the framework conditions, such as a lack of financial resources make its use even more difficult.
As remote sensing has a range of applications, the following recommendations for action can only provide an initial starting point for the successful application of remote sensing. It is based on experiences from development cooperation projects – but, given the high dependence on the context, it cannot and should not provide a blueprint for further action.
If possible, use free remote sensing data (e.g. Landsat or Sentinel satellite data) and/or derived products (e.g. land use maps) instead of re-analysing raw data. Many data portals provide global or regional products for various topics free of charge and with an ever-increasing spatial resolution. Their use minimises the costs for data procurement and analysis. The redundant development of tools or products is avoided.
Not all questions can be answered with these data portals and products. In some cases, the underlying procedures do not take adequate account of national definitions or methods. In addition, global products sometimes only contain generic information, whereas precise statements on crops or tree species are required.
But, in this case, the free data portals and applications – if they are adequately documented or software and codes are published – can provide a good starting point for the development of new products or monitoring systems, which are then adapted to the national context.
- In Germany, for example, Copernicus data are used to support the national monitoring of sustainability goals and the Sendai indicators at the administrative level. You can find out more here.
- Applications such as SEPAL provided by the FAO facilitate the use of remote sensing in forest monitoring and can be integrated into national monitoring systems in order to, for example, support “Measuring, Reporting, Verification“ (MRV) systems of the REDD+ mechanism. Find out more here.
As the procurement of certain remote sensing data can be associated with high costs, you should clarify the level of detail that you need to tackle your question before the start of the project. Freely available Landsat (NASA) or Sentinel (Copernicus) satellite data are often adequate. These missions achieve a ground resolution of up to 10m. This allows forests, larger watercourses and lakes or agricultural fields to be monitored.
However, this is often insufficient to monitor infrastructure, single tree felling activities or very small agricultural parcels due to the insufficient resolution of the recordings. In this case, high-resolution satellite data need to be purchased from commercial satellite operators. These days, such images have a ground resolution of up to 30cm. They show details, such as the transport infrastructure or parcel boundaries and are an ideal basis for mapping activities, among other things.
Images from aerial surveys or drones can have an even higher resolution. But the legal conditions (permits to fly) need to be taken into account when using these systems. In addition, aircraft and drones cover much smaller areas than satellites, meaning that their use is not suitable for national and extensive forest monitoring, for example. This is where satellites once again have the advantage.
Example of different spatial resolutions. Left: 0.40 metres, right: 10 metres, source: DigitalGlobe Open Data Program, CC BY 4.0, WorldView, 2018 (revised). Source: https://www.digitalglobe.com/ecosystem/open-data.
- First clarify whether government institutions in the partner countries have access to high-resolution satellite data and whether these can be shared and reused; pay attention to applicable licensing terms.
- Where appropriate, submit an application for free, high-resolution satellite images to the European Space Agency (ESA) – however, there is no guarantee and the assessment of applications can take several weeks.
- The market for available satellite data is large and complicated, as there are numerous specialist providers of high-resolution satellite images. In addition to the exclusive sale of satellite data, services such as the preparation of data, consulting and project cooperation are also offered. You can also consult the companies or the Copernicus network of experts in Germany.
The implementation of remote sensing in existing, official structures or processes to, for example, specify monitoring systems at a national level (MRV systems in the REDD+ context) has established itself in practice. This also includes participative processes in which data represent an important decision-making basis but where these data are incomplete.
The necessity of collecting comprehensive, consistent data, coupled with the low level of financial resources for on-site visits or the operation of in situ measuring systems can be a great incentive for the use of remote sensing.
Targeted cooperation with institutions in the partner country with existing know-how and technical expertise in remote sensing facilitates the acceptance and the implementation of remote sensing in projects or their implementation in existing processes.
- Create stakeholder maps: Before starting the project, create an overview of all the relevant stakeholders, e.g. universities or companies with geodata expertise and projects that can take over the analysis of satellite data.
- Use existing expertise in Germany
Know-how and professional networking
Establish know-how at the professional level, raise awareness at the decision-making level
Acceptance of the use of remote sensing is based on an underlying understanding of the potential applications as well as knowledge of the underlying methods (transparency). Maps and geodata are not just information, they can also be a political instrument, whether in the selection (or removal) of the information or the interpretation of the information.
In addition to providing data and applications it is therefore important that all potential users are also able to use the information. As a result, the qualification, particularly of government representatives, should be covered and the specialist and decision-making level should be consulted.
Remote sensing data should primarily be processed and validated by the counterpart in the project country. So be sure to check and systematically build on the existing expertise in the partner organisations. The institutional anchoring of this expertise in authorities can be supported by external consultants or remote sensing companies. After the end of the project, a key focus is on the long-term financing of local experts so that they remain at the relevant authorities.
- Integrate cost/benefit calculations and case studies: To further increase the acceptance of the use of remote sensing, case studies with a tried-and-tested application from other regions can be consulted in the initial project phase (examples are provided in the digital toolkit). The visual presentation of information can have a permanent impact at the decision-maker and management level in raising awareness amongst employees, counterparts and project partners. Besides case studies, cost/benefit calculations can be used, particularly if the outcome of the project is the regular use of remote sensing. A few starting points are provided here: remote sensing provides geographically consistent data and information in large and inaccessible regions and reduces the necessity for expensive and time-consuming on-site visits.
- Further information: A study of the European Commission highlights how political decisions are specifically supported by remote sensing. Other case studies and inspirational projects in the context of Agenda 2030 are available here and on the remote sensing site of the Federal Ministry of Economic Cooperation and Development’s (BMZ) digitalisation toolkit.
- International networking: Professional networking at an international level should be actively supported, e.g. by participation in specialist conferences and workshops (expand expertise) as well as by participating in international initiatives (actively shape). The intergovernmental Group on Earth Observations (GEO) established in 2005 is an example of this. GEO has set up a number of professional initiatives, which address existing gaps in the need for information, especially of global programmes in the areas of forest protection (GEO Global Forest Observation Initiative GFOI), food security (GEO Global Agricultural Monitoring Initiative GEOGLAM), land degradation neutrality (GEO LDN Initiative) and the conservation of biodiversity (Global Biodiversity Observation Network GEO-BON). The opportunities for involvement are based on professional networking as well as on the possibility of learning internationally agreed methods and procedures, actively cooperating in their development, integrating national requirements and influencing the discussion. This allows methods and data to be harmonised internationally, which ultimately enables more coherent and consistent information from remote sensing data to be provided at the international level (e.g. for Sendai and SDG indicators).
Established decision-making processes, which derive information and ultimately decisions from remote sensing data, are often lacking. To create knowledge and acceptance for methods based on remote sensing and enable their use for reporting purposes, both the specialist as well as the decision-making levels need to be consulted.
Transparent and clear communication of possible errors and limitations of the products (accuracy analyses), which can also be understood by non-experts, establishes trust and increases acceptance of remote sensing.
Open communication of the methods used that can be understood by non-experts and validity of the data are important. The validation of remote sensing products should be incorporated in the project design and carried out by local experts who are ideally anchored in the relevant authorities and ministries, or by nationally accredited institutes. This can be supported by external consultants.
- Remote sensing products are validated with the help of in situ data or other data sources. Use internationally established methods and guidelines to make remote sensing products more consistent and comparable at an international level.
- Information on the topic of accuracy analyses and the key principles of recording in situ data is provided, for example, in a comprehensive guideline published by the FAO or in the REDDcompass.
Another aspect is the volume of data to be processed: while a few satellite images can certainly be analysed on a desktop PC, some issues require the analysis of numerous satellite images in large regions – in this case, the data volume can no longer be managed on private PCs. Cloud computing can provide an alternative.
Cloud computing refers to the general provision and use of IT infrastructure, such as memory, processing power or software and access to remote sensing data over the internet. This is based on the idea of providing this IT infrastructure via a computer network so that it does not have to be installed on local computers.
The benefits of cloud computing are that no separate hardware resources are required to store the huge data volumes and use the application software provided via cloud service. This makes hard drives and USB flash drives obsolete and there is no need to modify or upgrade in-house hardware. Another benefit is the possibility of accessing the relevant cloud service at any time and with different devices and saving the costs for IT specialists.
Disadvantages of cloud services include the handling of often sensitive, national data. The use of cloud services involves a dependence on a service provider (outsourcing of IT skills). The quality of the internet connection can also be a critical point, as cloud services can only be used online. In addition, data protection is an important topic, as the data, for example, are stored on a server in the USA, by Google or Amazon for instance, which do not necessarily comply with the relevant national data protection provisions.
- A prominent example of cloud computing with satellite data is Google Earth Engine (GEE) or SEPAL administered by the FAO. This allows large amounts of data to be processed online. The technology is already in use in countless applications.
- Besides GEE, there are also the European DIAS platforms and the Copernicus Data and Exploitation Platform – Germany (CODE-DE) developed for official users, among others. Benefits compared to GEE: The data are stored on European servers, which can increase acceptance amongst government users.
- Open data cubes (ODC) are open source solutions for accessing, administering and analysing large volumes of earth observation data, especially time series of satellite images. ODCs, such as Digital Earth Africa, enable partner countries to set up their own cloud solutions.
Remote sensing data can be used in project evaluation to derive measurable indicators to review impact hypotheses in order to quantify the intended impacts, e.g. reforestation and erosion protection measures or the increase in agricultural productivity. A key benefit of remote sensing in this context is that free remote sensing systems provide an adequate spatial resolution for a large number of evaluative issues (up to 10m) and that freely accessible archive data stretching back several decades are available. Many applications, web tools or even nationally implemented monitoring systems provide information that is potentially relevant for project evaluation.
However, these data do not always reflect the defined impact indicators. If satellite data have to be re-analysed for impact monitoring, in situ data must be taken into account when designing the project or the possibilities of using existing in situ reference data need to be clarified.
Remote sensing can provide valuable information even before the start of a project, e.g. as an objective data pool in the concept development phase. The analysis of remote sensing data allows the areas of intervention to be specified, e.g. by mapping hotspots of deforestation or degraded land. In conjunction with other data, statements can be made on the possible suitability of certain locations in GIS, for instance for reforestation or agricultural use (site analyses).
However, depending on the issues covered, in situ data may be required and local stakeholders and projects that use remote sensing may already exist. The topic of remote sensing should therefore already be raised in review missions in order to determine the specific added value for a project (who is the target group, what are the associated requirements, how can remote sensing specifically support the work). To do so, review missions could be supported by external consultants with remote sensing expertise.
The quantitative interpretation of remote sensing data without in situ reference data is not possible, as remote sensing information merely describes a status, the causal relationships can only be explained with background knowledge. As a result, in situ data are required for many issues and should be collected by local partners. Training sessions may be necessary in some circumstances. Service providers specialising in remote sensing or research institutions can provide support in this respect.
In situ data are primarily used to calibrate algorithms and models that derive information from the data. For example, in situ data may be information on land use measured on-site with a GPS (crops, forest cover). The signatures from the satellite image time series are, for example, compared against the actual cultivation situation observed in the field. This comparison is performed by an algorithm that learns from the assignment of the signature and observation and then uses the signature to determine the crop for fields for which no observation exists. In situ data are also used to validate remote sensing products, i.e. to assess how accurately the maps reflect reality. In the aforementioned example, in situ data are used to better understand how exactly different crops differ from one another in a map.
Depending on the specific issues, additional information is required, such as observation data from air-, sea- and ground-based sensors as well as reference and additional data, such as river level measurements, whose collection sometimes involves or has involved a great logistical effort. As a result, make sure that you clarify whether in situ data are required, the specific issue and the type of data in the project concept phase. Then explore the possible access to existing in situ data or the framework conditions for their collection as part of review missions. Local partner institutions should be responsible for procuring in situ data and the subsequent analysis of the remote sensing data.
Gain an overview of potential applications, inspirational projects and specific applications of remote sensing in the following topic areas: