1 Introduction

Demands for risk monitoring and risk-informed decision-making are expressed by the Sendai Framework (United Nations 2015). Similarly, the needs for integrative assessments, including basic infrastructure, urban features, and disaster risk, are expressed in the New Urban Agenda (UN/HABITAT 2017). However, while disaster risk studies have covered major cities and national capitals, studies on smaller cities or peri-urban areas are lacking widely (Birkmann et al. 2016).

Peri-urban areas are often among the fastest-growing urban areas when close to a major city or capital like Nairobi. The complexity of geospatial growth is analysed in the course of history, over the colonial period before 1963, the national phase up to 1980 and beyond, wherein peri-urban growth is bound by societal processes and politics of land access and use (Ndegeah 2015). Research on peri-urban growth for cities such as Nairobi has a long history and covers aspects of planning response to low-income settlers (Memon 1982), for example. Land-use change analysis shows that Nairobi’s built-up area grows by around 1.5 km2 per year (Mundia 2017). Growth rates as extracted by remote sensing data show a growth between 8.4 and 17.2 percent of the spatial area in Nairobi between 1995 and 2015 (Katyambo and Ngigi 2017), or between 2.91 and 15.63 km2 per year between 1988 and 2014 (Mundia 2017). Other geospatial studies have differentiated growth rates of land-use area per urban, agriculture, water, and other sectors between 1986 and 2010 (Mubea et al. 2014). Population growth rates for Nairobi are 4.6% between 1948 and 1962, and then 12% and 1969; later on, growth rates decreased, also in other major urban centres in Kenya, to around 5% (Mundia 2017). Urban growth also expands the ratio of built-up areas potentially exposed to natural hazards such as floods or droughts. Monitoring of disaster risk needs to capture static pictures of risk and include the dynamics to better understand risk and resilience development and identify pathways to govern urban areas towards better resilience in coping with shocks and disasters (Simonovic and Peck 2013). This notion of the importance of dynamic risk assessments is expressed by recent research on aspects of the transformation of cities (Pelling et al. 2015), pathways in coping with climate change (Gibson et al. 2016), or transition (Solecki et al. 2017). However, this leads to a demand for data available for long periods to analyse changes.

Limitations of static risk assessments have long been acknowledged (King 2001), but few studies provide change assessments of disaster risk. This is often due to the lack of data for comparison or awareness about existing sources of data. While this is a challenge existing worldwide, availability and access to data are a major constraint for many countries in Africa (Osuteye et al. 2017). It seems paramount to improve data availability for risk assessments, and this paper points to data sources not used widely yet.

There is a wealth of studies utilizing remote sensing data and urban area classification approaches (Bhatta 2010; Patino and Duque 2013). Multi-temporal and time-series analyses have increasingly been applied in this area of urban and land-use mapping (Acevedo and Masuoka 1997; Li et al. 2018), or peri-urban growth (Shaw and Das 2018). Mapping transformation using remote sensing increasingly starts applying multi-spatial-resolution change detection (Zhang et al. 2016). However, many approaches in remote sensing studies focus on the technical advancements and are less applied for the operational mapping of urban transformation yet. And while urban growth studies are common (Bagan and Yamagata 2012; Masek et al. 2000; Song et al. 2016), fewer studies combine it with a spatial detection of exposure to natural hazards, especially over a range over more than 20–30 years in comparison. Even fewer studies have used satellite data before the 1970s, such as the declassified satellite data of the USA (Day et al. 1998). The USA has disclosed declassified former espionage imagery from the 1960s onwards that are hardly used for risk assessments so far (Fekete 2020). Declassified satellite data from the CORONA, GAMBIT, or HEXAGON series have been applied rather selectively and not very often yet, to fields such as archaeology (Ur 2003), generation of digital elevation models (Altmaier and Kany 2002), glacier monitoring (Bolch et al. 2008; Narama et al. 2010), land-use (Tappan et al. 2000), or urban growth (Masek et al. 2000; Stewart et al. 2004; Cecchini et al. 2019; Hepcan et al. 2013).

There is an increasing demand for multi-modal, multi-spectral, and multi-temporal remote sensing approaches (Brito and Quintanilha 2012; Qiming 2011; Stiller et al. 2021), and making use of a combination of aerial and satellite images can offer opportunities for that.

Aerial imagery and old declassified former espionage reconnaissance satellite imagery are major data sources to capture previous land cover back to the 1940s. Kenya has been covered by UK Royal Airforce aerial imagery campaigns due to the former colonial history. Access to such data is still restricted; it is on data servers not known much outside remote sensing communities. A lot of this data is open access, but sometimes, fees have to be paid; an aerial image from the UK National Collection of Aerial Photography starts at 16.5 £ plus tax, plus publication licenses, if needed. The CORONA satellite images from the US Geological Survey cost 35$, when not already freely available on their Earth Explorer website platform.

Although the costs are relatively low, as compared to recent high-resolution satellite imagery, it can still mean a restriction for researchers or local institutions. Another data access challenge is finding hazard data. This is the case for industrialized countries as well as for many other countries too.

Geospatial assessments are used in Kenya for monitoring land-use change or property rights (Koeva et al. 2017). Aerial imagery is also used for monitoring the sustainable development of agriculture (Ekbom et al. 2001). Challenges in defining urban boundaries and overestimating urban areas by single criteria such as population density are analysed using aerial imagery (Potts 2017). Hazards analysed using aerial images cover many hazards, including erosion and contamination, but also deal with lake basin (Khan et al. 2011; Habib et al. 2009) or riverine floods (Olang and Fürst 2011), flash floods (Hoedjes et al. 2014), rainfall variability (Gamoyo et al. 2015), or vector-borne diseases concerning floods (Pope et al. 1992). The early warning seems to be a topic of special interest for the whole country (Hoedjes et al. 2014) or the Tana river basin (Otieno et al. 2019). The Tana river is covered by some studies (Leauthaud et al. 2013). But other areas are covered much less (Appendix). Existing land-use change assessments do not use data earlier than the mid-1970s when LANDSAT satellite imagery became used. For example, Nairobi has been analysed regarding land-use and land cover change for 1976–2000 (Mundia and Murayama 2010). However, systematic multi-temporal assessments of exposure of settlements to floods, droughts, or other hazards have not been conducted yet, only urban growth or exposure mapping for single static time periods. But there is a great potential applying flood exposure mapping for the assessment of multiple risks related to floods, such as flood risk indices (Hategekimana et al. 2018), health risks (Okaka and Odhiambo 2019), or early warning (Otieno et al. 2019).

Therefore, the objective of this paper is to expand existing land cover change assessments back to the 1940s using aerial imagery. This enables capturing urban growth outside the national capital, Nairobi, along urban rims and in expanding towns in Kenya. A transferrable mapping approach is demonstrated, and it is indicated which natural hazard data could be used to analyse urban growth into potential hazard zones over time. Natural hazards include river floods in this case, and exposure of settlement area to be flooded is the main focus here. Using heterogeneous remote sensing data from different satellite missions and with different sensors and resolutions, a temporal comparison is made possible that exceeds previous approaches. In addition, this approach can analyse data-poor areas and areas not covered yet by research on natural hazards and exposure, which may inspire similar research in other areas. The main aim of this article is to explore the potential about data sets existing in early aerial and satellite imagery and how to use them for disaster risk assessment purposes. It also shows how such data can be used for the built-up mapping in peri-urban areas of African cities. Another ambition is to investigate the link between built-up dynamics and flood susceptibility.

2 Method

Geospatial assessments are useful to cover large areas of land by one comparable data source. It needs to be coupled with ground-truthing and thereby often is used in combination with other modes of assessment, such as local field visits, statistical data, or others. However, geospatial data provides another form of multimodal characterization, which is the special issue's focus. This case is a multi-modal selection of greyscale aerial photographs, satellite-based greyscale imagery, multi-spectral recent satellite data, statistical data, and on-the-ground photos and field visits. Geospatial data are beneficial for analysing changes over time, using the same data source, or in combination with other modes of satellite, aerial remote sensing, ground-based imagery, or many other forms of data integrated into a geographic information system (GIS). The general methodology consists of a spatial assessment using QGIS to map urban areas and hazard zones. The first methodological step consists in the selection of the case study sites which is conducted in discussions with partners in Kenya. The main part is a mapping of urban areas and hazard zones using aerial and satellite imagery in combination with open-source GIS data. The areas derived from this assessment allow to calculate percentages of growth over several decades. The overall methodological process is shown in Fig. 1 and explained in more detail in 2.1.

Fig. 1
figure 1

Overview on the methodological steps

Full size image

2.1 Case study areas

The case study areas were selected in a co-design approach with local experts from Dedan Kimathi Technical University in Nyeri and Kenyatta University in Nairobi in joint workshops in Kenya on Sept 5, 2019, and online Feb 17, 2021. The two areas have been selected for demonstrating and testing the data availability and GIS assessment opportunities of a combination of old aerial, former satellite espionage, and recent satellite and OSM data. The case study areas have been selected, as one area is the peri-urban rim of Nairobi, the nation's capital and largest city (Fig. 2). The area selected in the North-East is where urban sprawl and growth occur to a great extent; it is Ruiru town in Nairobi East along the Thika Road. The other area selected also experiences urban growth, the city of Nyeri, located North of Nairobi, close to the Aberdare and Mt. Kenya mountain ranges. Thereby, both case studies represent examples of peri-urban areas growing in Kenya, but in different settings, one the national capital and mega city and the other a developing smaller city in a more rural setting. The local partners have selected them also, since they are familiar with them and can provide guidance during the study.

Fig. 2
figure 2

Selected case study cities, Nairobi and Nyeri, are within the most populated areas in Kenya (data: left tile: population density in 1989, data by WRI 2007, neighbouring countries by OSM 2021 data. Right tiles: GHSL population grid by (Schiavina et al. 2019), urban boundaries mapped using Sentinel-2 data)

Full size image

Satellite data useful for the land cover analysis and available for the study area between Nairobi and Nyeri in Kenya include LANDSAT coverage starting from 1973 (MSS), OrbView 3 starting from 2003, and Sentinel-2 data starting from 2015. Kenya is at the equator, and resolutions of digital elevation models are accessible in resolutions of about 250 m (Global Multi-resolution Terrain Elevation Data 2010—GMTED2010), and in about 30 m by the ALOS Global Digital Surface Model or the Shuttle Radar Topography Mission—SRTM. Satellite images on flood events have been searched at the Dartmouth flood observatory, but the images of two events recorded in the 2000–2008 Rapid Response Inundation Maps including Nyeri and Nairobi are not accessible anymore for the study area. The UNITAR/ UNOSAT map library contains data on two events from 2018 to 2019, but not from the case study area. The UNOSAT Flood Portal contains one event from 2008 in Kenya, but the event did not affect the case study areas. Earlier UNOSAT and other satellite maps are found by web search, but do not cover the case study areas. At the UN-SPIDER knowledge portal, no data set on floods and Kenya could be found. Within The International Charter Space and Major Disasters, six activations for Kenya could be found starting from the year 2000, five of them related to floods, but none of them from Nairobi or Nyeri.

Kenya has altitudes between zero and over 5,000 m; Nairobi is at 1,795 m and Nyeri at 1,762 m, situated along the Aberdare mountain ranges (Fig. 3). Mount Kenya with 5,199 m is just north of Nyeri and about 140 northeast from Nairobi. To the east, the topography decreases in altitude towards the coast.

Fig. 3
figure 3

SRTM digital terrain model of Kenya

Full size image

Both case study locations, Nairobi and Nyeri, experience major urban growth but are prone to natural hazards such as floods and droughts.

2.2 GIS analysis of land use using spatial imagery

The analysis is a GIS assessment using remote sensing data in combination with statistical data. QGIS is used as the GIS software since it is cost-free, open source, and enables the usage with different computer operating systems; so, it can be used by partners in Kenya and Germany and other countries, in teaching as well as in further research. The projection used is WGS84 UTM Zone 37 N, so metric analyses can be conducted.

First, urban growth is analysed by change detection using aerial images from 1948, declassified CORONA mission (Day et al. 1998), related data from 1967 to 1977, and Sentinel 2 data from 2020. The earliest available data and sufficient spatial resolution, coverage of the case study area and cloud coverage, and costs were selection criteria for the data (Table 1 provides details on remote sensing data ordered and used). The aerial images were ordered from the National Collection of Aerial Photography (NCAP) from the UK. The declassified CORONA, HEXAGON, and GAMBIT mission satellite images were obtained from the USGS Earth Explorer portal; three images were already available for free download. One additional image had been ordered. The OSM and Sentinel-2 data were openly available on OSM and EOX platforms.

Table 1 Aerial and satellite imagery used for this study
Full size table

The aerial photographs and CORONA satellite images are georectified based on OSM and Sentinel-2 data as base maps. Histogram stretch and contrast corrections have been applied. The following land-use classes have been mapped then; urban built-up areas, vegetation, and flood hazard areas. Urban built-up area features are manually extracted and digitized from the 1948 aerial photos and the CORONA satellite images at a viewing scale of 1:2,500 to 1:5,000 in QGIS. Sentinel-2 data are digitized at spatial scales between 1:5,000 and 1:10,000. To capture built-up areas, mainly houses are mapped since they are the easiest to identify by rectangular shapes as distinct urban areas. Houses are identified mainly by the rectangular shape of the roofs. In addition, roads and paved (/impervious/sealed) areas are mapped when connected to buildings but not gardens or other partly natural surfaces. Special care is taken along rivers to capture houses and urban features. Vegetation land cover is mapped especially for human-shaped irrigated land use when it is identifiable by rectangular features such as roads or borders to other vegetation.

For training purposes, 100 × 100 m grids are created (Fig. 4). An urban area is defined when adjacent urban features cover an area of at least 100 m2. Roads are only mapped when directly connected to buildings, up to 100 m to the buildings. Open spaces within a city are mapped as urban when surrounded by a built-up area. The mapped areas are then compared to recent land cover data from Sentinel-2 and Open Street Map (OSM) data to capture urban growth and land-use change, but also in some cases to check the plausibility of urban features still existing at the same spot.

Fig. 4
figure 4

Training areas to delineate urban from other land-use signature with 100 m grid. Red arrows indicate built-up area not included into the urban boundary

Full size image

Flood hazard features are analysed using water areas and bodies, river course, and other hydrological data available from different data providers (OSM, Regional Centre for Mapping of Resources for Development, SRTM, World Bank Data Catalogue, World Resources Institute). Since no flood hazard zonation maps are available and no DEM in sufficient resolution, a buffer zone is estimated around rivers in areas with rather even topography, crossing the city quarter of Nairobi North-East and Ruiru according to the average width of wetlands in data. It is also checked against a visual interpretation based on Sentinel-2 data from 2020. From the spatial data available as satellite imagery and auxiliary topographic maps, the signatures of vegetation and marshes around the rivers are visually identified. In the GIS, the distances of their perimeters around the rivers are measured and distances of 150–200 m are obtained. For further GIS assessment using buffers around the river lines, a 150 m distance is selected to remain more on the cautious side of estimations, rather than 200 m. The topography is checked based on digital elevation information available, and no steep terrain is detected that would demand for modifications of the buffers. At field visits in 2018 and 2019 at and around the Kenyatta university campus in this study site, it could be seen that the terrain is rather flat. Urban and hazard growth rates are then calculated based on the area sizes extracted from the aerial and satellite images.

3 Results

3.1 Nairobi: Ruiru

The mapping of the aerial image from 1948, the CORONA satellite data from 1977, and the Sentinel-2 data from 2021 reveals urban growth in north-eastern Nairobi (Fig. 5). The growth between 1948 and 1977 already is significant. It has to be noted that some of the settlement areas from 1948 as mapped from the aerial photograph are not mapped on the 1977 CORONA satellite image since the resolution is coarser. While some of the features, such as rectangular shapes that could be houses also, are visible on the 1977 CORONA image, they are not mapped when the signal is not clear enough to retain a consistent mapping approach.

Fig. 5
figure 5

Mapped area North-East of Nairobi, along Thika Road, at Ruiru town. Urban built-up area from 1948, 1977, and 2020

Full size image

The assessment in Nairobi reveals urban growth, as well as a growth of built-up area into potentially flood hazard-exposed zones, too (Fig. 6). To estimate flood exposure, buffer zones were created at 150 m width and have been checked against the topography as derived from a 30-m resolution SRTM model, as well as against existing wetland data sets. The terrain is rather even, and therefore, buffers extend to both sides of the river course quite evenly. However, this GIS assessment and the buffer provide only a first indication, which areas to analyse further with on-ground observation and finer-grained elevation models. Not all terrain within the buffer, including higher and lower terrain segments, will be flooded at all times.

Fig. 6
figure 6

Urban growth and 150-m buffer river zone at Nairobi Thika Road NE, at Ruiru town

Full size image

Using GIS, the area sizes of the built-up areas are calculated. The urban built-up area at the urban rim of Nairobi in the north-east in the subset used in Fig. 6 is calculated based on the imagery to an area of 0.016 km2 in 1948, 0.349 km2 in 1977 and 3.783 km2 in 2020. This is a growth rate of about 10.8% from 1977 to 2020 (Table 2). Ruiru town (red area in Fig. 6) has grown from area of 0.016 km2 in 1948 to 0.349 km2 in 1977, which is a growth rate of 26.5%. This growth is then followed by a slightly lower rate of 23.5% from 1977 to 2020, up to 3.783 km2 in 2020.

Table 2 Built-up and hazard area sizes and growth as calculated from the aerial and satellite images
Full size table

The results can be summarized that for Ruiru, the built-up area had grown 26-fold from 1948 to 1977 and 23-fold from 1977 to 2020. At the same time, the potentially flood exposed area had grown 1.7 times from 1948 to 1977 and over 100 times from 1977 to 2020. The exposed lengths of river embankments are provided which can be of interest for planning flood protection measures such as dikes. The analysis also reveals that Ruiru had grown much more than the adjacent South-Eastern part of the outskirts of Nairobi between 1977 and 2020. And the flood exposure of Ruiru had grown even much more by comparison to the built-up area. The growth rate of Nyeri between 1967 and 2020 is quite comparable to the area analysed in Nairobi NE/ Ruiru. However, since satellite data with sufficient resolution were not available from the same years, a direct comparison is not possible; it only can indicate similarities. Flood exposure also has increased in Nyeri, as the perimeter of the built-up area also has grown along the river Chanya.

Land-use changes are also observable within the Ruiru area as well as in Nyeri. In Ruiru, some areas had already been visibly transformed from natural environments to agricultural usage (Fig. 7). The map shows the aerial image of 1948 in the centre, with greyscales ranging from values 8 to 255. It overlays the CORONA satellite image underneath, which is cut to represent greyscale values from 4 to 239, to improve the signature contrast. Interestingly, many areas that were used intensively in 1948 (orange borders on the map) were not used in the same way in 1977. In 1948, the signature of the aerial image revealed darker colours and regular shapes of both roads crossing it and regular parallel linear structures as common for crops. In 1977, these areas revealed a speckled and heterogeneous reflection with lighter and mixed greyscales. On the other hand, some areas were intensively used in 1977, which had not been used in this way in 1948 (red borders on the map). Some other areas had been used both in 1948 and 1977, presumably for agriculture (not mapped).

Fig. 7
figure 7

Land-use changes in Ruiru from 1948 to 1977. (Data: see Table 1 for the aerial image and CORONA data)

Full size image

These land-use changes of some areas are important for mapping urban growth, too, since they show the landscape before settlement. In the southern part of the aerial image, one larger area is mapped (Fig. 8). It can be important to monitor such previous land usages for loss of biodiversity or for revealing settlements that have overgrown previous wetlands. It also shows up opportunities for drought monitoring and understanding long-term changes.

Fig. 8
figure 8

Development from intense agricultural use in 1948 to mixed-use in 1977 and the rather dense built-up area in 2020 in Ruiru. (Data: see Table 1 for the aerial image, CORONA and Sentinel-2 data)

Full size image

3.2 Nyeri

In Nyeri, a similar strong urban development took place between 1967 and 2020 (Fig. 9). Aerial images from before the 1960s could not be obtained, so no comparison to the 1940s was feasible. The location of Nyeri along the Chanya river indicates a potential flood risk, confirmed by the local experts from the Dedan Kimathi Technical University, who provided recent on-site video footage during the online workshops. Urban growth reveals an expansion of the potentially flood-exposed built-up areas. Of course, local topography must be observed, which is quite steep in this terrain and means that a river flood may not affect both sides equally.

Fig. 9
figure 9

Urban growth in Nyeri from 1967 to 2020

Full size image

In Nyeri, the terrain is steeper than in Nairobi–Ruiru. However, the 150-m buffer zone still fits quite well to the shape of the river Chanya valley, as derived from the STRM digital elevation model (DEM) (Fig. 10). But of course, due to the steep terrain, a much more accurate DEM data set and ground observation are necessary to evaluate the flood exposure risk. The SRTM data set exhibits some data processing flaws, as visible by large rectangular shapes of the slopes in the upper left part of Fig. 10. The slope degrees range from 0 to 48, as described in the figure legend, and high degrees of inclination (Fig. 10) also point out a potential exposure risk to landslides and erosion. The circles in Fig. 10 indicate areas for further observation on-ground, where bridges and buildings could be at risk of flooding or mass movement.

Fig. 10
figure 10

Urban boundaries of Nyeri in 2020 overlayed on digital elevation model (SRTM), with Chanya river 150-m buffer zone

Full size image

4 Discussion

The case study analysis reveals the usage of GIS in combination with remote sensing data for monitoring land-use change over several decades. Specifically, the results consist of digital maps that are based on data with restricted spatial and temporal resolution. These maps can help to guide planers and disaster risk managers to identify exposure risks within a broader regional overview. Such data can complement more detailed risk assessments on the ground. Spatial data have the advantage to cover large regions with the same sensors and derived information signatures. It can also enable comparisons between time phases and document existing as well as historical land-use situations. Significantly, this study shows how aerial imagery and older declassified satellite data can expand the time range covered.

4.1 Data and methodology constraints

It has become evident that many constraints exist, such as in spatial resolution and temporal availability. Satellite and aerial imagery missions in the 1980s and before were limited in temporal coverage, spatial resolution and also, not all areas have been covered in a country. Typically, larger urban areas or other areas of strategic military interest are often mapped primarily (Day et al. 1998). For the case studies, it was more difficult obtaining images for Nyeri than for Nairobi. But even for Nairobi, older missions have mainly mapped the urban area at that time. At urban rims such as Ruiru, the availability of imagery was constrained. For example, the high-resolution image from 1964 did not cover these outskirts of Nairobi anymore.

The manual mapping is constrained in precision and handling for larger areas and related amount of work. Automatization development of semi-automated and automated machine learning unsupervised land-use classifications can enable certain improvements in the work-flows and precision, while having their own constraints. However, since the aerial and satellite imagery is greyscaled and extracted from analogous film, manual interpretation and mapping still proved useful. Some attempts with the 2020 Sentinel-2 in using automated classifications of land-use or extraction of built-up areas revealed problems; the pattern of land-use and settlements in Kenya differs much from other areas in the USA, for example. This is due to different dense vegetation cover in Kenya and more speckled settlement types. Many smaller houses and shacks along roads and rural areas make it difficult to separate their signature from other features. Some huts in rural areas growing into peri-urban areas also have a rather rounded roof and stand intermixed with other features. Regular rectangular road layouts are often missing, and paved roads or open places usually make identifying built-up areas easier. The manual mapping approach proved to be more precise than earlier machine learning automated approaches compared with a publicly available data set (WRI 2007). But even up-to-date automatic extraction supported by artificial intelligence is constrained by data availability of Landsat data, ranging back globally to 1975, and its respective spatial resolution (Pesaresi et al. 2019; Schiavina et al. 2019). The Global Human Settlement Layer provides data for Kenya for 1975, 1990, 2000 and 2015 in 250 m resolution, or combined in a multi-temporal data set up to 30 m (Florczyk et al. 2019; Freire et al. 2016). For 2017 to 2018, 10 m spatial resolution is also available for the built environment and 2 m for 2015 data. Comparing the machine learning approaches with manual mapping, the manually extracted results still are more precise in this spatial resolution (Fig. 11). In addition, Google Open Buildings data were compared. The data are derived with help of artificial intelligence. Unfortunately, many mismatches and errors were identified in the two case study areas, such as parking lots mistaken for buildings, and other objects that are easy to identify visually by comparing it to Sentinel-2 data or other. But of course, the manual mapping is also constrained by original data resolution and mapping skills.

Fig. 11
figure 11

Manually mapped built-up areas in 2020 from Sentinel-2 data, as compared with the Nyeri built-up area in 1989 (left; data by WRI 2007) and 1975 and 2015 for Ruiru (right; data by (Schiavina et al. 2019)) extracted automatically by machine learning approaches

Full size image

But of course, manual mapping and visual interpretation also provide many challenges; for example, land-use change specification of vegetation types or identification of geomorphological features (Glass 2013). Multi-spectral imagery is much more apt to differentiate and identify vegetation types as well as urban features and separate built-up from other land-use types (Patino and Duque 2013; Wentz et al. 2014). The application of NDVI and built-up area assessments was analysed in this study for Nairobi using Sentinel-2 data, but was not as precise and successful as in other areas such as the USA (Fekete and Priesmeier 2021). It was more difficult, because the urban sprawl around Nairobi provides a specific mix of vegetation and land use for farming within those areas, small huts and shacks. In comparison to the same data for the USA, fewer rectangular shapes of roads and urban layout exist, which hampers urban feature identification. Remote sensing offers a range of further options for multi-spectral data or data in high spatial resolution, such as usage of thermal sensors (Voogt and Oke 2003), night-time light urban feature extraction (Wentz et al. 2014; Zhao et al. 2019), object-based classification (Blaschke 2010; Kucharczyk et al. 2020), LiDAR data (Yan et al. 2015; Wedajo 2017), or radar data (Soergel 2010). This is promising for recent satellite data but has limitations when applied to greyscale older satellite images with often medium or rather coarse spatial resolution.

4.2 Multimodal enrichment opportunities

This sub-section discusses options for future assessments, to advance the results and study shown in this article. It is presented using a new data set example and relates to specific research directions in GIS assessments that are useful to advance the study further.

The presented GIS assessment can be extended from land-use change mapping to enrichment by additional data and interpretations. For example, when comparing with recent OSM data, features such as previous river beds that now have been overgrown by settlement areas could be added to OSM (Fig. 12).

Fig. 12
figure 12

Identification of previous river beds (1977) within recent settlement areas (2020) in Nairobi Githurai (Data: CORONA image left and OSM data right)

Full size image

Other options would be to add further sources of VGI, such as photos of locations after flood events or Twitter messages (Taylor 2016). Another option is to add community efforts in preparing flood response measures, as mapped in Kibera, a slum in Nairobi, removing garbage, widening rain channels, diverting floods with sandbags, etc. (Nelson 2016). Another important possibility is adding vulnerability and resilience information, for example, about the informal settlements which cover many areas in Nairobi. Certain slums in the city centre have been covered by studies (Corburn and Karanja 2014), but peri-urban areas lack coverage so far. The potential of remote sensing data, however, has been demonstrated to capture such informal settlements and could show an important method to cover such areas (Kraff et al. 2020; Taubenböck et al. 2018) where data availability and local accessibility often are a challenge, and ethical aspects must be observed, too. Other features of urban composition can be analysed by spatial data, too, such as central business districts that carry another type of economic risk (Taubenböck et al. 2013). But next to social vulnerability, dependency of urban areas from critical or baseline infrastructure is an upcoming topic, for example, as demanded by the United Nations (United Nations 2015). Urban populations depend on piped water, roads, food, and many other urban services that should be integrated into holistic risk assessments (Ledant 2013). But also, the natural environment needs to be included in such holistic risk assessments. For example, the ecosystem services, the risk, the value of nature-based solutions, and investments into sustainable land use (Vogl et al. 2017). And while this article has shown applications only to flood hazards, it could also be applied to erosion and landslide risk.

5 Conclusions

The article demonstrates how urban growth into flood-prone areas can be mapped using aerial and satellite images over several decades. Using GIS, the built-up area in Ruiru, Nairobi, had grown 20-fold from 1948 to 1977. At the same time, the potentially flood exposed area had grown 7 times from 1948 to 1977. The analysis shows that Ruiru had grown much more than adjacent outskirts of Nairobi between 1977 and 2020. The growth rate of Nyeri between 1967 and 2020 is quite similar to the selected Nairobi case study area in the north-east. Such results can provide an estimation for urban planners and disaster risk management institutions on increasing risk and exposure. The satellite and aerial images can also help to point out former land coverage such as river beds or wetlands that can pose risks such as risks related to groundwater or heavy rain.

As an outlook, current international bodies and concepts such as the Sendai Framework demand more risk-informed decision-making (United Nations 2015). The need has grown for data, monitoring, and data visualization for decision-makers and the public. Social inequalities, vulnerabilities, and countries in the global South are a focus area due to natural hazard events and fewer resources to cope with damages and losses (Warner and Zakieldeen 2012; Wrathall et al. 2015). However, large areas, especially in the global South, are largely uncovered for even basic risk monitoring of urban growth or spatial exposure to common natural hazards such as river floods. Data scarcity adds only to the problem and underlines the knowledge gaps persisting (Osuteye et al. 2017). Overall, this may point towards a distorted understanding of the risk of societies, where the disaster risk could be less known or understood in some countries less covered by data and respective risk assessments. Hitherto rather unexplored data and respective data repositories such as aerial imagery captured by former colonial governments or espionage reconnaissance missions could provide key innovations to expand the current limited understanding of the dynamic spatial development of risk. The article also shows caveats of limited availability and usefulness of old data when constrained by accessibility, spatial, and spectral resolution. Visual mapping is time-consuming and limited to detecting certain types of natural hazards such as floods or landslides, droughts, or wildfires. But it is a good starting point for risk mapping and expanding the application of remote sensing data.

Such temporal monitoring of exposure to natural hazards can help inform not only a static risk but also the dynamics of risk components such as urban expanse and settlement on former riverine areas over time. This can also support the current research trend on transformation in the context of climate change (Gibson et al. 2016; Pelling et al. 2015) about the opportunities existing for measuring changes using remote sensing data (Kraff et al. 2020). Of course, many more aspects need to be added to capture risks, such as information about vulnerabilities, capabilities, and resilience (Cutter et al. 2008). And for transformation, fundamental changes such as regime changes need to be added to capture what is currently understood as transformation (IPCC 2012, 2019; Solecki et al. 2017). However, land-use change is a fundamental footprint of human activity and the result of many decisions taken in combination, such as economic, legal, and environmental decisions. It is often difficult to capture how to avoid risks in context to on-going transformations and one example could be to better guide “build back better” programmes (Gibson et al. 2016; Kennedy et al. 2008) on how to avoid reconstruction in areas where settlement has overgrown former river beds. Mapping spatial exposure transformation can therefore be understood as a tool helping to document and understand a larger transformation. In context to climate change, more documentation of the hazard trends and human interaction with natural hazard-prone areas is necessary, especially in countries underresearched and hampered by data scarcity.