Person | TU Delft Repository

D.J. de Villiers

5 records found

Authored

Enhancing water access monitoring through mapping multi-source usage and disaggregated geographic inequalities with machine learning and surveys

Journal article (2023) - J. Geleijnse, J. Geleijnse, M.M. Rutten, D.J. de Villiers, James Tayebwa Bamwenda, E. Abraham

Monitoring safe water access in developing countries relies primarily on household health survey and census data. These surveys are often incomplete: they tend to focus on the primary water source only, are spatially coarse, and usually happen every 5-10 years, during which significant changes can happen in urbanisation and infrastructure provision, especially in sub Saharan Africa. In this work, we present a data-driven approach that utilises and compliments survey based data of water access, to provide context-specific and disaggregated monitoring. The level of access to improved water and sanitation has been shown to vary with geographical inequalities related to the availability of water resources and terrain, population density and socio-economic determinants such as income and education. We use such data and successfully predict the level of water access in areas for which data is lacking, providing spatially explicit and community level monitoring possibilities for mapping geographical inequalities in access. This is showcased by applying three machine learning models that use such geographical data to predict the number of presences of water access points of eight different access types across Uganda, with a 1km by 1km grid resolution. Two Multi-Layer-Perceptron (MLP) models and a Maximum Entropy (MaxEnt) model are developed and compared, where the former are shown to consistently outperform the latter. The best performing Neural Network model achieved a True Positive Rate of 0.89 and a False Positive Rate of 0.24, compared to 0.85 and 0.46 respectively for the MaxEnt model. The models improve on previous work on water point modeling through the use of neural networks, in addition to introducing the True Positive - and False Positive Rate as better evaluation metrics to also assess the MaxEnt model. We also present a scaling method to move from predicting only the relative probability of water point presences, to predicting the absolute number of presences. To challenge both the model results and the more standard health surveys, a new household level survey is carried out in Bushenyi, a mid-sized town in the South-West of Uganda, asking specifically about the multitude of water sources. On average Bushenyi households reported to use 1.9 water sources. The survey further showed that the actual presence of a source, does not always imply that it is used. Therefore it is no option to rely solely on models for water access monitoring. For this, household surveys remain necessary but should be extended with questions on the multiple sources that are used by households.@en

Something fishy going on?

Evaluating the Poisson hypothesis for rainfall estimation using intervalometers: results from an experiment in Tanzania

Journal article (2021) - D.J. de Villiers, M.A. Schleiss, J.A.E. ten Veldhuis, R.W. Hut, N.C. van de Giesen

A new type of rainfall sensor (the intervalometer), which counts the arrival of raindrops at a piezo electric element, is implemented during the Tanzanian monsoon season alongside tipping bucket rain gauges and an impact disdrometer. The aim is to test the validity of the Poisson hypothesis underlying the estimation of rainfall rates using an experimentally determined raindrop size distribution parameterisation based on Marshall and Palmer (1948)'s exponential one. These parameterisations are defined independently of the scale of observation and therefore implicitly assume that rainfall is a homogeneous Poisson process. The results show that 28.3 % of the total intervalometer observed rainfall patches can reasonably be considered Poisson distributed and that the main reasons for Poisson deviations of the remaining 71.7 % are non-compliance with the stationarity criterion (45.9 %), the presence of correlations between drop counts (7.0 %), particularly at higher arrival rates (ρa>500 m−2s−1), and failing a χ2 goodness-of-fit test for a Poisson distribution (17.7 %). Our results show that whilst the Poisson hypothesis is likely not strictly true for rainfall that contributes most to the total rainfall amount, it is quite useful in practice and may hold under certain rainfall conditions. The parameterisation that uses an experimentally determined power law relation between N0 and rainfall rate results in the best estimates of rainfall amount compared to co-located tipping bucket measurements. Despite the non-compliance with the Poisson hypothesis, estimates of total rainfall amount over the entire observational period derived from disdrometer drop counts are within 4 % of co-located tipping bucket measurements. Intervalometer estimates of total rainfall amount overestimate the co-located tipping bucket measurement by 12 %. The intervalometer principle shows potential for use as a rainfall measurement instrument.@en

Intervalometer (Slight Return)

Poster (2019) - N.C. van de Giesen, D.J. de Villiers, R.W. Hut, M.A. Schleiss, J.A.E. ten Veldhuis

An intervalometer is an extremely simple measurement device that measures the intervals between the impact of raindrops on a surface. In our case, we used a piezo-electric element put in a small 3D-printed holder. When a raindrop hits the surface, a voltage is generated that is d ...

Contributed

A completer picture of domestic water access and consumption

Integrating machine learning models and survey information

Master thesis (2022) - J. Geleijnse, E. Abraham, M.M. Rutten, D.J. de Villiers, D. van Halem

The United Nations (UN) Sustainable Development Goal (SDG) \#6 reads that by 2030 universal and equitable access to safe and affordable drinking water is achieved for all In order to achieve this goal, proper and complete monitoring, capturing all the facets of safe water access, is essential. In this thesis it is argued that the current monitoring techniques come with limitations and subsequently solutions are presented to come to a more complete picture of water access.

Monitoring safe water access happens primarily through household health surveys. These surveys are often incomplete, not covering entire nations, focus on only the primary water source and are often spatially aggregated for privacy reasons. Besides, health surveys almost never include questions on consumed water volumes while that is an important indicator for proper hygiene (WELL, 1998), and something that, at the same time, should be in balance with the natural available water resources. Next to this survey based monitoring, there is the Water Point Data Exchange (WPDx) that monitors safe access by providing a platform at which the exact location and type of water access points (such as boreholes, springs, etc.) are registered. This does give more insight into the presence and usage of a variety of sources, but also the WPDx is often incomplete: not covering entire nations.

In this thesis we present a dual methodology that gap-fills the incompleteness of the WPDx database through modeling and in parallel, researches the complex local dynamics of water access, the variety of water sources used by households and the relationships between access and water consumption by means of a household survey.

By improving a machine learning biological species modeling technique (called MaxEnt), successful predictions on the number of presences of eight different water access types across Uganda were made, also into areas that have little presence in the WPDx data. It was found that population density, precipitation, elevation, poverty and groundwater storage are important indicators for the (non)presence of water access points.

Next to modeling, a survey campaign was executed in Bushenyi-Ishaka municipality, a mid-sized town in the South West of Uganda comprising a mixture of both urban and rural areas. This was done in collaboration with Makerere University (Kampala). The survey results showed that water consumption increases with education and wealth, but also with higher number of water point presences predicted by the model. It was also found that households in Bushenyi make use of an average of two different water sources on a regular basis and often express preference for sources off premises compared to on premises (piped) for both cost and perceived quality reasons.

Lastly, modi operandi were suggested for the results to improve water access such as prioritising areas with poor(est) water access and investing in rainwater harvesting, infrastructure and education.

Sustainability of groundwater resources in Kumasi, Ghana

Estimating potential groundwater recharge using a water-balance approach and remote-sensing data in Google Earth Engine

Master thesis (2021) - E. Fernandes Potter, M.M. Rutten, D. van Halem, I. Monney, D.J. de Villiers

Groundwater use has seen a significant increase in the rapidly urbanising city of Kumasi, Ghana, due to its reliability, general good quality, and low-cost development. Conversion of vegetated to urban land along with the challenges of growing groundwater abstraction has put the groundwater system at risk. This study aims to assess how the urbanisation trends in Kumasi affect the availability of groundwater resources. Sustainability is investigated by comparing (multi-)annual groundwater withdrawals to long-term average annual replenishment. Groundwater recharge is estimated in Google Earth Engine using a water-balance approach and remote-sensing datasets to bridge the hydrological data gap. Runoff is determined using the Soil Conservation Service Curve-Number (SCS-CN) method requiring soil and land-cover maps. Land-cover maps are created for 1986, 2013, and 2020 using Sentinel-2 and Landsat 8 surface reflectance products. Evapotranspiration is derived from the MODIS evapotranspiration product and precipitation is retrieved from CHIRPS. Results indicate that groundwater recharge decreased by 80% from 1986 to 2020 (124Mm3 to 28.9Mm3) attributed to a loss of permeable land (63 percentage point decrease) and intensification of urbanisation. Domestic and non-domestic groundwater consumption in 2020 is estimated to be 28.4Mm3 and 2.7Mm3 respectively. For 2020, long-term average groundwater recharge (28.9Mm3) is less than the annual abstraction (31.1Mm3) indicating current urbanisation trends are unsustainable for future groundwater availability. Under a "business as usual" scenario, population growth presents the largest challenge for the future by causing an estimated four-fold increase in groundwater consumption by 2050. Climate change and land-cover changes under the same scenario may reduce groundwater recharge by 10% and 55% respectively. This requires the implementation of policies to properly manage the groundwater resource, such as promoting low-impact development, monitoring groundwater use, and monitoring changes in the groundwater system. This study should be seen as a preliminary investigation into the components affecting groundwater sustainability. Further research is needed to assess the exact state of the groundwater system and its response to future challenges.

https://code.earthengine.google.com/b4ebac49f28c93a633eb7bd4806f409b (groundwater recharge)
https://code.earthengine.google.com/ad1a0a88240e75e3dd727a6a350edb18 (land cover changes)