When rain falls on Earth, the water starts moving and flowing downhill through sewers and rivers as runoff. Runoff is extremely important to recharge surface water bodies and groundwater. Furthermore, runoff changes the landscape by action of erosion. It is an integral part of the water cycle (Earth Science Data Systems 2021). 

Problems related to runoff occur when the water flows in an uncontrollable way on land and cannot be absorbed anymore: this excess of water pours outside rivers, estuaries, and other freshwater areas (National Geographic n.d.). Runoff can come from a natural source such as snowmelt, erosion and weather, or from anthropic activity such as pipes or irrigation (National Geographic n.d.). If you would like to read more about snowmelt monitoring, we recommend reading this Space4Water article: a hidden secret that becomes water: monitoring Patagonia Glacier Retreat. It is also soil dependent, with a much higher overflow risk in urban areas where the land is covered by impervious surfaces (U.S. Geological Survey 2018).

Runoff can damage the environment in many ways: when coming from a polluted source or coming across a polluted land (driveways, agricultural fields), it can transport contaminants into local streams by washing them from the surface (US EPA 2015; Earth Science Data Systems 2021). The transported pollutants include nutrients, pesticides, petroleum chemicals, metals and even antibiotics (National Geographic n.d.). In freshwater, these pollutants can harm an entire ecosystem, cause harmful algal blooms or dead zones and in some cases they can make water ecosystems (like wetlands or mangroves) out of use. According to the United Nations, in the past century, 85 percent of natural wetlands have been lost, whereas artificial ones have been built (United Nations n.d.; Earth Science Data Systems 2023). Wetlands allow us to remove pollutants from runoffs, and hence represent a mitigation tool for clean water and sanitation (Vymazal and Březinová 2015). If you would like to read more about harmful algal bloom, we recommend reading this Space4Water article; if you would like to read more about Wetlands conservation, read this Space4Water article.

As much as runoff is an environmental threat, it is also a threat to the economy, menacing food production and freshwater availability in a world with a growing population. To limit water quality loss caused by polluted runoff, reliable monitoring is needed at national and international levels. Monitoring runoff can provide important information on the sources of pollution of a water body (Earth Science Data Systems 2021). Failing to detect polluted runoff can be a threat to achieving Sustainable Development Goal 6.3.2: proportion of bodies of water with good ambient water quality. To reliably monitor runoff accurate data is required. Earth Observation (EO) can complement local ground data and support data acquisition (UN Water 2018) and informed decision making at a national level (If you would like to read more about the importance of space technology in quantifying freshwater availability globally, read this Space4Water article).

Several scholars argue (Ghosh, Jaiswal, and Ali 2021; Hong et al. 2007; Huo et al. 2021) it is never easy to monitor runoff in remote mountain areas and highlands with bad weather conditions, complex topography and undeveloped economies. Moreover, Hong et al. (2007) argue that the Northern Hemisphere has higher numbers of sampling, which provide more accurate data, penalising Southern countries with less sampling and less precise data (Hong et al. 2007).  If you would like to read more about data collection and citizen science, we recommend reading this Space4Water article: Crowdsourcing and Citizen Science data for water resources management. The lack of data can impact risk management and, in our case, runoff monitoring.

Nonetheless, to successfully model runoff, many combinations of technologies can be operated of which a few of them are briefly described in the next chapter. For example, Huo and al (2021) monitored runoff in a Tibetan basin via passive microwave data (data related to the temperature and moisture properties of the surface) collected from the Scanning Multichannel Microwave Radiometer (SMMR), the Special Sensor Microwave/Imager (SSM/I), Special Sensor Microwave Scanning Radiometer-Earth/Sounder (SSMIS) and the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) (Huo et al. 2021; Canada 2008). Socio-economic data can also be used to complement remote sensing research. In a study made in a humid tropical watershed in Brantas, Indonesia, Wiwoho and Asturi defined the study area with available socio-economic data and were able to identify four non-socioeconomic factors that could impact runoff: rainfall, elevation, slope and green vegetation fraction. These four factors were studied by the authors because “the degrading conditions in Brantas are apparently linked to the multifaceted pressures on the watershed”. An Artificial Neural Network (ANN) was then used to quantify the sensitivity of these different factors to runoff (Wiwoho and Astuti 2022).

Using remote sensing technologies to monitor runoff

Satellites cannot measure runoff directly, but they allow us to collect information on land surface properties such as Land-Use-Land-Cover (LULC), and to forecast precipitation amounts and intensity, and humidity (If you would like to read more about monitoring precipitation from Space, read this Space4Water article). These data are then input into land surface data to model runoff and estimate their impact on water quality (Earth Science Data Systems 2023). Many countries use EO data to manage potential risks related to runoff and plan projects allowing for sustainable runoff management. Examples include Armenia, Timor-Leste, Indonesia to name just a few. 

The Armenian government received support by the Asian Development Bank (ADB) to map snow covered mountains to model runoff for planning and operating winter tourism infrastructures. ADB also provided support to a project on the transport sector in Timor-Leste. They used EO data to model climate expected effects on runoff along the roads. The Government of Indonesia used EO data for flood forecasting as part of their flood risk management in river basins in the Banten and the Maluku Provinces. They assessed LULC and the topography within watersheds to identify important areas for protection and restoration as well as to reduce extreme runoff. Challenges they faced in the assessment include great variations in data accuracy and outdated landcover data (Locsin and Aschbacher, n.d.). 
There is a variety of ways to monitor runoff with EO data. The land surface properties studied by the National Aeronautics and Space Administration (NASA) include soil texture, topography, evapotranspiration, and leaf area index, as it influences runoff (Earth Science Data Systems 2021) (If you would like to read more about monitoring hydrological changes from space in a sparse gauged basin, read this Space4Water article). Digital Elevation Models and LULC are important EO data products to derive information on runoff or even erosion potential (Locsin and Aschbacher, n.d.). With this data known, it is possible to simulate runoff.

NRCS-CN method coupled with NAPI

Figure 1: Global NRCS runoff CN map derived from U.S. Department of Agriculture hydrological soil groups and land cover classification for fair hydrological conditions (Hong et al. 2007)
Figure 1: Global NRCS runoff CN map derived from U.S. Department of Agriculture hydrological soil groups and land cover classification for fair hydrological conditions (Hong et al. 2007)


To simulate runoff, the NASA approached the thematic with a simplistic method using satellite rainfall estimation. They calculated and compared three sets of Global Runoff Data according to climate by using the concept of Normalized Antecedent Precipitation Index (NAPI) and the Natural Resources Conservation Service runoff - Curve Number method (NRCS-CN). 

The CN represents the infiltration rate of the soil: the higher the CN, the lower the infiltration rate is, the less permeable the soil is and the higher the risk of runoff is (HEC-RAS Hydraulic Reference Manual 2023). The CN is a parameter that allows to approximate the soil type, land cover and Antecedent Moisture Condition (AMC) by calculating the infiltration (potential retention of water) and runoff generated by rainfall accumulation (Hong et al. 2007). NRCS-CN helps generating maps by estimating surface runoff as a function of precipitation and the CN (figure 1). 

Coupling NAPI with NRCS-CN method allows predicting runoff according to the 5 previous days of rainfall and soils moisture data in the basin. This in turn allows to forecast runoff for the catchment in case no runoff data is available. Despite a given uncertainty, the forecast can be considered useful for approximating runoff for the globe and medium to large river basins (Ghosh, Jaiswal, and Ali 2021). 

Artificial Neural Network based simulations of the rainfall runoff process with mixed sources of remote sensing input data

Figure 2: Scheme of an Artificial Neural Network (TIBCO n.d.)
Figure 2: Scheme of an Artificial Neural Network (TIBCO n.d.)


The use of a CN method can be challenging due to mismanagement and inappropriate agricultural practices, and artificial intelligence can be preferred. To simulate the rainfall-runoff process in a watershed in Iran, Gholami and Sahour (2022) used  Artificial Neural Networks (ANNs) on data from field sampling plots in conjunction with rainfall and hydrometric data. Hydrometric data are standardized water quantity data and information. ANNs are machine learning algorithms that are widely used in hydrological studies and have proved to be a capable tool for rainfall-runoff modelling because of their high precision and time efficiency compared to more traditional methods (Senthil Kumar et al. 2005). The author’s network uses a combination of remote sensing input data including rainfall time series, soil AMC, initial loss (the amount of rainfall required to wet a catchment before runoff starts), and time to peak of the basin to determine the runoff time series. ANN models are capable to represent extremes values of rainfall-runoff, allowing to predict floods and low flows. But, according to the authors, “the major drawback of ANN-based models is their limitation in disclosing the relationship between input and output [runoff time series] due to the black-box nature of models” (Gholami and Sahour 2022). 

Using socio-economic data to trace runoff

When EO data is lacking, socioeconomic data can be used to trace back past runoff. A study made in Southern China retraced 30 years of socioeconomic development and showed a correlation between the increase of resource consumption and ecological interference, with the intensification of ecological problems such as droughts and floods. The authors of the study further concluded that the wealthiest countries had the least erosion while poorest tropical countries were the most susceptible to high levels of soil erosion (Shuxia et al. 2021). 


Runoff is a threat to the environment, human health and to the economy and can impact the SDG 6 directly. EO data can provide help to model runoffs to set up risk management measures. Even if runoff cannot be measured directly, mapping runoff using EO data such as LULC and weather data (AMC, rainfall intensity and frequency) is possible. Calculation of indexes or the use of machine learning using remote sensing data combined with in-situ data allow to simulate runoff. These methods are used according to the climate of the study area, but one sure thing is that there are multiple methods that can be used to monitor runoff in different climate areas, according to data availability and reliability.  


Canada, Natural Resources. 2008. “Microwave Remote Sensing.” Natural Resources Canada. January 29, 2008. https://natural-resources.canada.ca/maps-tools-and-publications/satelli….

Earth Science Data Systems, NASA. 2021. “Water Quality Data Pathfinder - Find Data.” Data Pathfinder. Earthdata. Earth Science Data Systems, NASA. September 15, 2021. https://www.earthdata.nasa.gov/learn/pathfinders/water-quality-data-pat….

2023. “Agriculture and Water Management Data Pathfinder.” Data Pathfinder. Earthdata. Earth Science Data Systems, NASA. March 4, 2023. https://www.earthdata.nasa.gov/learn/pathfinders/agricultural-and-water….

Gholami, Vahid, and Hossein Sahour. 2022. “Simulation of Rainfall-Runoff Process Using an Artificial Neural Network (ANN) and Field Plots Data.” Theoretical and Applied Climatology 147 (1): 87–98. https://doi.org/10.1007/s00704-021-03817-4.

Ghosh, Narayan C., Rahul Kumar Jaiswal, and Shakir Ali. 2021. “Normalized Antecedent Precipitation Index Based Model for Prediction of Runoff from Un-Gauged Catchments.” Water Resources Management 35 (4): 1211–30. https://doi.org/10.1007/s11269-021-02775-w.

HEC-RAS Hydraulic Reference Manual. 2023. “Curve Number.” 2023. https://www.hec.usace.army.mil/confluence/rasdocs/ras1dtechref/latest/o….

Hong, Yang, Robert F. Adler, Faisal Hossain, Scott Curtis, and George J. Huffman. 2007. “A First Approach to Global Runoff Simulation Using Satellite Rainfall Estimation: TECHNICAL NOTE.” Water Resources Research 43 (8). https://doi.org/10.1029/2006WR005739.

Huo, Junjun, Xing Qu, Dejun Zhu, Zhe Yuan, and Ziyue Zeng. 2021. “Runoff Monitoring in the Lhasa River Basin Using Passive Microwave Data | Elsevier Enhanced Reader.” 2021. https://doi.org/10.1016/j.jag.2021.102486.

Locsin, Ma Carmela, and Josef Aschbacher. n.d. “Earth Observation for a Transforming Asia and Pacific.”

National Geographic. n.d. “Runoff.” Accessed March 21, 2023. https://education.nationalgeographic.org/resource/runoff.

Senthil Kumar, A. R., K. P. Sudheer, S. K. Jain, and P. K. Agarwal. 2005. “Rainfall-Runoff Modelling Using Artificial Neural Networks: Comparison of Network Types.” Hydrological Processes 19 (6): 1277–91. https://doi.org/10.1002/hyp.5581.

Shuxia, Yu, Xie Chenyu, Jinsong Zhao, Wang Zhen, Wang Ling, and Shi Zhihua. 2021. “Socioeconomic Development Mitigates Runoff and Sediment Yields in a Subtropical Agricultural Watershed in Southern China.” Environmental Research Letters 16 (November): 12. https://doi.org/10.1088/1748-9326/abdd5a.

TIBCO. n.d. “What Is a Neural Network?” TIBCO Software. Accessed April 18, 2023. https://www.tibco.com/reference-center/what-is-a-neural-network.

UN Water. 2018. Sustainable Development Goal 6: Synthesis Report 2018 on Water and Sanitation. United Nations Publications. New York, New York, United States of America: United Nations.

United Nations. n.d. “Goal 6 | Department of Economic and Social Affairs.” Accessed March 21, 2023. https://sdgs.un.org/goals/goal6.

US EPA, REG 01. 2015. “Soak Up the Rain: What’s the Problem?” Collections and Lists. August 20, 2015. https://www.epa.gov/soakuptherain/soak-rain-whats-problem.
U.S. Geological Survey. 2018. “Runoff: Surface and Overland Water Runoff.” 2018. https://www.usgs.gov/special-topics/water-science-school/science/runoff….

Vymazal, Jan, and Tereza Březinová. 2015. “The Use of Constructed Wetlands for Removal of Pesticides from Agricultural Runoff and Drainage: A Review.” Environment International 75 (February): 11–20. https://doi.org/10.1016/j.envint.2014.10.026.

Wiwoho, Bagus Setiabudi, and Ike Sari Astuti. 2022. “Runoff Observation in a Tropical Brantas Watershed as Observed from Long-Term Globally Available TerraClimate Data 2001–2020.” Geoenvironmental Disasters 9 (1): 12. https://doi.org/10.1186/s40677-022-00214-5.