The research, led by Milad Rajaei with Usman Safder, Sarah Cotterill, and Recep Kaan Dereli, presents a city-scale framework that simulates water demand in near real time by explicitly tracking how people move through a city. The work builds on SIMDEUM, a well-established stochastic end-use model originally developed in the Netherlands, which represents water consumption as a stream of random pulses—each toilet flush, shower, tap use, or dishwasher cycle—with timing, duration, and flow rate drawn from probability distributions derived from empirical observations of occupant behaviour. SIMDEUM has proven remarkably capable of reproducing realistic household demand patterns at high temporal resolution, and it has been extended to offices, hotels, nursing homes, and other non-residential buildings by dividing each building into functional rooms with their own appliances and users. But the model has always carried a fundamental weakness: it assumes occupancy is either static or averaged, which in dynamic urban environments can become the dominant source of error.

The importance of occupancy is not in doubt. Sensitivity analyses of stochastic residential demand models have found Spearman’s rank correlation coefficients between occupancy and both peak and average demand ranging from 0.99 to 0.995—an almost perfect relationship. Field experiments reinforce the point: researchers who installed flush counters on 119 toilets across seven university campus buildings demonstrated a strong, direct link between toilet water use and the number of people present. The COVID-19 pandemic made the consequences of ignoring this relationship vividly clear, as commuting collapsed, workplaces emptied, and hygiene practices intensified, producing higher residential demand alongside sharply reduced commercial consumption. Models grounded in static occupancy assumptions simply could not see these shifts coming.

The researchers’ insight was to recognise that the data needed to model dynamic occupancy already exists in aggregate form. Google Popular Times indicators describe how busy non-residential locations are at any given moment, derived from aggregated and anonymised smartphone location data, expressed on a relative scale from 0 to 100 compared with a location’s typical peak activity. The framework begins by collecting these signals at five-minute intervals through automated web scraping for every non-residential building in a study area. Where live data are available and pass quality checks, they are used directly; where they are not, the model descends through a careful hierarchy of fallbacks—historical average patterns for the same day of the week, then representative occupancy profiles derived from K-means clustering of tens of thousands of profiles collected nationwide, then literature-based profiles from U.S. Department of Energy reference buildings, adjusted with local correction factors for seasonal effects such as school terms and hotel occupancy statistics.

Converting relative busyness into absolute occupant numbers requires a further step: each Popular Times value is multiplied by the estimated capacity of the building, calculated by dividing floor area by occupancy load factors taken from building design and fire safety guidelines, and scaled to reflect normal operation rather than maximum permitted crowding. Data quality proved to be a genuine challenge. Across a one-month collection period in September 2024, 58 percent of live samples were classified as invalid under the study’s quality-control rules, which flagged suspicious sudden drops in occupancy that persisted briefly before abruptly returning to normal—patterns unlikely to represent real activity. Days with insufficient valid data were replaced wholesale with historical averages, while shorter gaps were filled by linear interpolation. The prevalence of anomalies, particularly at low-traffic locations, underscores that crowdsourced occupancy data is useful but demands rigorous preprocessing.

The most conceptually ambitious element of the framework is its treatment of residential occupancy, for which no direct crowdsourced signal exists. Rather than relying on census averages, the model infers where people are at home by tracking population movements. Drawing on two classic theories of human mobility—Zipf’s gravity model, which holds that movement likelihood rises with population and falls with distance, and Stouffer’s intervening opportunities model, which assumes people choose the nearest destination that satisfies their needs—the framework constructs a trip probability matrix at each time step. When non-residential occupancy rises, the corresponding number of people is drawn probabilistically from residential areas weighted by their populations and the distribution of nearby opportunities; when occupancy falls, people return to their original home areas. A tourist population, calibrated from national accommodation and tourism statistics, handles movements associated with hotels and nightlife, while a separate commuter population accounts for people travelling into the study area from outside.

These time-varying occupancy estimates then feed directly into a modified SIMDEUM model running at one-minute resolution in MATLAB. At each time step, the probability of a water-use event for each end-use is calculated from the occupancy, the per-person frequency of use, and a diurnal timing factor reflecting behavioural routines. A random draw is compared with this probability to decide whether an event occurs, and if so, its flow rate and duration determine the volume consumed, with each end-use temporarily locked during an event to prevent overlap. For occupancy-dependent end-uses such as toilet flushing and hand washing, the occupancy term drives the calculation; for scheduled activities such as office cleaning, occupancy is effectively set aside so that only frequency and timing matter.

Applied to Sligo, a coastal town of roughly 20,000 people in northwest Ireland, the framework simulated an entire month of city-scale demand. Residential consumption came out at approximately 129 litres per person per day—closely matching the metered benchmark of about 312 litres per household per day reported for Sligo—and the simulated end-use breakdown, with toilets accounting for 28 percent of consumption, showers 24 percent, and kitchen taps 21 percent, differed by no more than two percentage points from published values for Irish households. The temporal patterns behaved as one would expect: a pronounced morning peak between 6:00 and 9:00 a.m. on weekdays driven by showering and breakfast routines, a delayed peak on weekends, a midday dip as residents left for work or school, and an evening recovery as people returned home. Non-residential demand told equally coherent stories—restaurants showed sharp peaks aligned with mealtimes, food retail displayed the steadier profile of continuous cleaning and toilet use, and office buildings peaked at the start of the working day before collapsing after closure.

The study is candid about its limitations. The default SIMDEUM parameters derive from Dutch household data and may not transfer cleanly to Irish conditions—simulated restaurant water use of 6.2 litres per square metre per day diverged from the 2.48 litres reported by Irish Water for comparable commercial premises, a discrepancy the authors attribute primarily to uncalibrated appliance frequencies, durations, and flow rates. The conversion of relative busyness into absolute occupant counts also requires independent validation against footfall sensors or building occupancy systems. Nevertheless, the results demonstrate something genuinely significant: a scalable, transferable route to high-resolution water demand modelling that requires no dense smart metering infrastructure whatsoever. By resolving demand at the level of individual end-uses across entire cities, the framework opens the door to demand-responsive pumping schedules, energy-aware operation of distribution networks, and scenario testing for planners—capabilities that could meaningfully reduce the energy intensity and emissions of the urban water cycle, one flush at a time.

Subject of Research: A real-time, city-scale water demand modelling framework that integrates urban mobility data from Google Popular Times with a stochastic end-use water demand model.

Article Title: A framework for real-time water demand modelling at city scale based on urban mobility

Article References: Rajaei, M., Safder, U., Cotterill, S., & Dereli, R. K. (2026). A framework for real-time water demand modelling at city scale based on urban mobility. Energy Reports, 16, Article 109700. https://doi.org/10.1016/j.egyr.2026.109700

Image Credits: AI Generated

DOI: 10.1016/j.egyr.2026.109700

Keywords: water demand modelling, urban mobility, Google Popular Times, SIMDEUM, smart meters, occupancy, water-energy nexus, end-use model, stochastic simulation, urban water systems, Sligo Ireland, energy efficiency

The research, conducted by Ibrahim L. Kadigi of Mbeya University of Science and Technology, draws on microdata from the 2019/20 National Sample Census of Agriculture, a comprehensive survey covering more than 5,000 rice-growing households across Mainland Tanzania, with the earlier 2007/08 census used to normalize and stabilize yield comparisons. Rather than relying on simple averages or regression models, the study deployed a nonparametric stochastic simulation framework. Using Latin Hypercube Sampling and a Multivariate Empirical distribution, the analysis generated 500 simulated yield draws for each of 22 unique combinations of seed type and agroecological zone, producing 11,000 simulated yields in total. This Monte Carlo approach captures the full probability distribution of farm yields, including variance, skewness, and extreme outcomes, offering a far richer picture of production risk than conventional mean-based comparisons.

The methodological core of the study is the so-called stoplight analysis, which ranks the probability that farms exceed, fall between, or drop below defined productivity thresholds. Under Scenario A, the researchers set a lower cut-off of 1.5 tonnes per hectare and an upper benchmark of 3.0 tonnes per hectare, roughly the national target range. Under Scenario B, aligned with global standards, the thresholds rise to 2.0 and 4.5 tonnes per hectare. Because average rice yields in Tanzania hover around 2.4 tonnes per hectare, well below the global average of about 4.7 tonnes and far short of the 5 to 7 tonnes achievable with better management, these thresholds effectively frame the gap between ordinary and exceptional performance.

The results are unambiguous. At the national level, 29 percent of farms using improved seeds exceeded the 3.0 tonnes per hectare benchmark, compared with only 20 percent of local-seed farms, a 45 percent increase in the likelihood of high yields. Meanwhile, the share of farms falling below the 1.5 tonnes per hectare floor dropped from 38 percent among local-seed users to 30 percent among improved-seed users. On the mainland, the advantage widened further: 36 percent of improved-seed farms surpassed 3.0 tonnes per hectare versus just 21 percent of local-seed farms, a 71 percent relative improvement, and the proportion of very low-yielding farms fell from 36 percent to 23 percent. In Zanzibar, where overall productivity is lower, improved seeds still nearly doubled the probability of exceeding the benchmark, from 11 percent to 19 percent, and cut the low-yield share from 52 percent to 41 percent.

When the more demanding global thresholds were applied, the absolute probabilities of top-tier performance shrank, but the relative advantage of improved seeds grew even more pronounced. Nationally, the share of farms exceeding 4.5 tonnes per hectare doubled from 6 percent among local-seed users to 12 percent among improved-seed users, while the proportion of farms yielding less than 2.0 tonnes per hectare fell from 61 percent to 50 percent. On the mainland, 15 percent of improved-seed farms cleared the 4.5 tonnes per hectare bar, more than double the 6 percent of local-seed farms, and the low-yield share collapsed from 60 percent to 41 percent. In Zanzibar, improved-seed users were more than three times as likely as local-seed users to reach the global benchmark, and the share of extremely low-yielding farms declined from 79 percent to 64 percent. The pattern, the study emphasizes, is that improved seeds shift the entire yield distribution upward rather than simply nudging averages.

The agroecological breakdown adds crucial nuance. Tanzania’s rice landscapes span nine distinct zones, from the semi-arid Central Zone with erratic rainfall of 400 to 700 millimeters, to the fertile, well-watered Southern Highlands receiving 900 to 1,400 millimeters annually. In the Northern Highlands, improved seeds raised the probability of exceeding 3.0 tonnes per hectare from 50 to 55 percent and essentially eliminated the risk of yields below 1.5 tonnes per hectare, dropping it from 4 percent to zero. In the Southern Highlands, high-yield probabilities climbed from 34 to 52 percent while low-yield risk was nearly halved from 28 to 15 percent. Even in structurally disadvantaged zones, improved seeds delivered measurable gains: in the Southern Zone, the share of farms exceeding the benchmark rose from a dismal 2 percent to 20 percent, and in the Eastern Zone, high-yield probabilities increased from 14 to 27 percent while low-yield risk fell from 32 to 21 percent.

Yet the study is candid about limits. In the Southern and Western zones, more than half of farms remained in the low-productivity band even with improved seeds, and under the higher global thresholds, low-yield probabilities in some zones climbed to 68 to 79 percent despite adoption. The Western Zone even showed a slightly higher probability of very poor yields among improved-seed users under the strictest scenario, at 79 percent versus 73 percent for local seeds. These findings signal that seed quality alone cannot overcome deeper constraints such as soil acidity, poor drainage, limited irrigation, and inadequate extension coverage. The authors argue that seed subsidies and input programs should therefore not be applied uniformly nationwide but calibrated to zone-specific bottlenecks, pairing improved seeds with irrigation expansion, water harvesting, soil fertility restoration, lime subsidies, and drainage improvement where those constraints dominate.

The adoption picture itself reveals a stark institutional divide. While fewer than 5 percent of mainland farmers use improved rice seeds, adoption in Zanzibar reaches nearly 30 percent, a difference the study attributes partly to targeted seed subsidy schemes, strengthened last-mile distribution through shehia-level cooperatives, and public-sector seed multiplication programs on the islands. The low mainland uptake reflects weak seed systems, high transaction costs, market inefficiencies, limited extension services, and affordability barriers, compounded by the fact that many smallholders rely on informal, recycled seeds with declining genetic purity and germination performance. The census data also cannot distinguish certified improved seeds from quality-declared or farmer-saved improved varieties, a limitation the author notes may cause the estimated productivity differences to understate or overstate the true performance of certified material.

Validation of the simulation framework was rigorous. Observed and simulated yield distributions overlapped closely across all four major farmer groups, with means, standard deviations, coefficients of variation, and distribution shapes aligning almost perfectly. For instance, mainland improved-seed farms showed an observed and simulated mean of 2.77 tonnes per hectare with identical standard deviations of 1.39 and coefficients of variation of roughly 50 percent. No outliers were removed, since rare high-yield observations plausibly reflect practices such as the System of Rice Intensification, and the nonparametric framework deliberately retains tail outcomes to represent the full spectrum of production risk. The close correspondence reflects correct model specification rather than overfitting, the study explains, because simulated draws are generated directly from empirical distributions.

The policy implications extend well beyond Tanzania’s borders. Rice consumption in Sub-Saharan Africa has surged from roughly 13 kilograms per capita annually in the 1970s to more than 30 kilograms today, and Africa imported over 16 million metric tonnes of rice valued at more than USD 6 billion in 2022 alone. Tanzania, despite cultivating more than 1.1 million hectares, still imports between 70,000 and 120,000 metric tonnes annually. By quantifying where improved seeds most effectively lift yield distributions and where complementary investments are essential, the study offers actionable evidence for the Agricultural Sector Development Programme II, the National Rice Development Strategy II, and the Tanzania Seed Sector Development Strategy 2030, while supporting progress toward Sustainable Development Goals on poverty, hunger, and climate action. In high-potential zones, the priority is scaling, mechanization, and market integration; in marginal zones, seeds must ride alongside water, soil, and knowledge investments. Improved seeds, the evidence shows, are a necessary lever for Tanzania’s rice transformation, but never a sufficient one on their own.

Subject of Research: The effect of improved rice seed adoption on rice productivity and production risk across agroecological zones in Tanzania, using national census microdata and nonparametric stochastic simulation.

Article Title: Improved seeds and rice productivity in Tanzania: policy insights from national sample census data

Article References: Kadigi, I. L. (2026). Improved seeds and rice productivity in Tanzania: policy insights from national sample census data. BMC Agriculture, 2(1), Article 35. https://doi.org/10.1186/s44399-026-00042-0

Image Credits: AI Generated

DOI: 10.1186/s44399-026-00042-0

Keywords: improved seeds, rice productivity, Tanzania, agroecological zones, stochastic simulation, National Sample Census of Agriculture, food security, smallholder farmers, yield thresholds, seed systems, ASDP-II, NRDS-II