The study, published in Frontiers in Sustainable Food Systems, focuses on Jiangxi as a major agricultural province and one of China’s first National Ecological Civilization Pilot Zones. Its indicator system covers resource conservation and efficient use, production-site environmental governance, agricultural ecological restoration, green industrial development, and green technological support. Together, these dimensions contain 18 secondary indicators, including farmland retention, irrigation efficiency, fertilizer use efficiency, pesticide intensity, crop-straw utilization, forest coverage, soil and water conservation, rural income, agricultural mechanization and scientific-technological progress. The design was tailored to Jiangxi’s mountainous and hilly terrain, abundant forests, extensive inland waterways and exposure to high temperatures, droughts and floods.

To calculate the historical index, the researchers range-standardized official data from 2010 through 2025 and assigned weights using the entropy weight method, which reflects the amount of temporal variation and information differentiation in each indicator. The resulting dimension weights were 29.42% for agricultural ecological restoration, 23.28% for green industrial development, 22.97% for production-site environmental governance, 16.29% for resource conservation and utilization, and 8.04% for green technological support. These weights describe statistical variation within the observed sample, not causal influence or policy importance. A robustness check using the CRITIC method produced the same ranking of the five dimensions, while indicator-level weights had a Spearman correlation of 0.91, with p<0.01.

The historical trajectory divided into three periods. During an initial exploration phase from 2010 to 2015, the index increased from 11.37 to 32.63, averaging 4.25 points per year. Growth accelerated between 2016 and 2020, when the score rose from 32.63 to 64.84, an average of 6.44 points annually. This fastest phase coincided with Jiangxi’s designation as a National Ecological Civilization Pilot Zone and the implementation of fertilizer and pesticide reduction programs, high-standard farmland construction and agricultural-waste utilization measures. From 2021 to 2025, the index advanced from 64.84 to 80.51, averaging 3.13 points per year as the transition moved from major input reductions toward more systematic improvement.

Several measured changes accompanied the faster middle phase. From 2016 to 2020, chemical fertilizer application intensity declined from 25.1 to 22.6 kilograms per mu, while pesticide application intensity fell from 0.82 to 0.61 kilograms per mu. Over the same interval, comprehensive livestock and poultry manure utilization increased from 70.3% to 85.2%, and crop-straw utilization rose from 82.1% to 89.7%. Over 2015–2025, fertilizer application intensity decreased from 26.8 to 21.3 kilograms per mu, while fertilizer-use efficiency increased from 33.2% to 41.5%. These indicators contributed to the modeled index, but the observational analysis does not establish that any one policy alone caused the composite change.

The forecasting method combined the entropy-weighted historical assessment with separate GM(1,1) grey prediction models for each of the 18 indicators rather than modeling only the final composite. For each nonnegative series, the first-order accumulated generating operation reduces short-term volatility and supports estimation of an exponential trend. Separate fitting windows were selected for the individual indicators, and the forecast values were then combined using the historical weights. The researchers retained the 2010–2025 minimum and maximum values as normalization bounds and restricted projected indicator values to that interval. All fitted models had mean absolute percentage errors below 10%, meeting what the authors describe as the Grade I accuracy standard for grey prediction.

Under this baseline, the projected composite score was 83.97 in 2026, 84.91 in 2027, 85.02 in 2028, 85.14 in 2029 and 86.97 in 2030. That is an increase of 6.46 points from the 2025 value, or about 1.29 points per year when measured across the full 2025–2030 interval; the average within 2026–2030 alone is approximately 0.75 points annually. At the dimension level, the model places faster growth in green technological support and green industrial development, slower growth in resource conservation and production-site governance, and steady improvement in ecological restoration. The increasing importance of mechanization in the forecast is associated with expected expansion of machinery suited to hilly farming.

The projection’s principal uncertainty lies in the conditions it assumes. Jiangxi’s 2024 comprehensive mechanization rate for crop cultivation, sowing and harvesting was 71.2%, nearly five percentage points below the national average, and machinery levels were lower in cash crops, facilities, livestock and aquaculture. Farmers operating less than 2 mu represented more than 60% of the total, limiting their capacity to purchase and maintain specialized equipment. At the same time, agricultural non-point source pollution control is becoming more costly as fertilizer and pesticide use becomes harder to reduce, while the frequency of heat, drought and flood events could disrupt trends in yields, input demand and infrastructure use. These are structural uncertainties that the historical-series models do not explicitly simulate.

Jiangxi’s certified agricultural sector provides a commercial context for the forecast. The combined number of products under the “Three Certifications and One Geographical Indication” framework rose from 3,363 in 2020 to 8,496 in 2025, while green-food products increased from 942 to 3,742 and organic products from 2,317 to 4,219. High-quality distinctive agricultural products increased from seven to 430, and green-food production bases expanded from 8.543 million to 9.294 million mu. The study notes that certification categories can overlap, creating a measurement limitation, and that growth in certification does not guarantee a market premium or higher farmer returns. It concludes that the projected index is best read as a reference trajectory for identifying likely progress and policy gaps, not as a certain future outcome or a substitute for direct assessment of environmental, economic and social effects.

Subject of Research: Entropy-weighted measurement and grey-model forecasting of agricultural green development in Jiangxi Province

Article Title: Construction of an agricultural green development indicator system and trend projection for the 15th five-year plan period in Jiangxi Province

Article References: Dai, T., Yu, W., Li, Y., & Liu, F. (2026). Construction of an agricultural green development indicator system and trend projection for the 15th five-year plan period in Jiangxi Province. Frontiers in Sustainable Food Systems, 10, Article 1942701. https://doi.org/10.3389/fsufs.2026.1942701

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1942701

Keywords: Jiangxi Province, agricultural green development, green agriculture, GM(1,1) model, entropy weighting, indicator system, food security, agricultural mechanization, environmental governance, trend projection, Construction, agricultural

The team assembled a retrospective longitudinal dataset spanning fourteen quarters, from the last quarter of 2020 through the first quarter of 2024, yielding 518 state-quarter observations. That granularity matters. Most previous analyses of Nigerian road safety have relied on annual aggregates or single-state case studies, which can obscure the seasonal rhythms and regional disparities that drive risk. By working at the state-quarter level, the researchers could track how crash burden shifted over time and across jurisdictions, and could calculate three key indicators for each state: the fatality rate, defined as deaths per crash; the casualty rate; and a severity index that captures how dangerous a typical collision becomes once it occurs.

The headline numbers tell a story of crisis followed by cautious relief. Reported crashes climbed to a peak of 13,656 in 2022, a year that also claimed 6,456 lives on Nigerian roads. In 2023, the crash count fell by 22.3 percent, a decline that will strike many readers as genuinely good news. But the study’s most sobering finding is what did not change. The fatality rate per crash, the probability that a collision produces a death, held steady between 0.47 and 0.48 deaths per crash from 2021 through 2023. In other words, fewer crashes happened, but each crash remained just as lethal. The decline in total deaths tracked the decline in crash numbers almost mechanically, suggesting that whatever reduced crash frequency did nothing to make the crashes that did occur more survivable.

Geographically, the analysis reveals a striking decoupling of crash frequency from crash lethality. The Federal Capital Territory recorded the highest number of crashes of any jurisdiction, 4,852 over the study period, yet its fatality rate was a comparatively low 0.224, meaning fewer than one in four of its crashes proved fatal. Contrast that with northern states such as Sokoto, Katsina and Kaduna, which logged fewer crashes in absolute terms but suffered substantially higher death rates per crash. A collision on a highway through Sokoto is, on the evidence of this dataset, roughly twice as likely to kill someone as a collision in the capital. The authors attribute this pattern to differences in road conditions, vehicle standards, speed profiles and the speed and quality of emergency medical response, factors that determine whether a crash becomes a statistic or a tragedy.

Population adjustment reshuffles the league table entirely. When crashes and fatalities are expressed per 100,000 residents, states that appeared moderate in raw counts emerge as carrying disproportionate burdens, while some populous states with high raw numbers turn out to have unremarkable per-capita rates. This distinction is not academic hair-splitting. Road safety budgets are finite, and a state with a modest crash count but a small population and a high fatality rate may deserve far more attention per naira spent than a sprawling urban center whose crash numbers simply reflect the volume of traffic. The study’s per-capita calculations make clear that part of the apparent geographic concentration of crashes is an artifact of where people live rather than where driving is genuinely most dangerous.

What causes these crashes? The researchers drew on the causal factors recorded by Federal Road Safety Corps officers at crash scenes, and two categories dominated: speed violations and vehicle-related defects. This is consistent with a long line of road safety research showing that kinetic energy is the fundamental killer in road trauma. The relationship between impact speed and fatality risk is steeply nonlinear; a collision at 100 kilometers per hour releases far more than twice the destructive energy of one at 50, and the human body’s tolerance for sudden deceleration has hard physiological limits. Vehicle-related issues, including brake failure, tire blowouts and faulty lighting, compound the problem, particularly on long-haul routes where commercial vehicles travel enormous distances under economic pressure to move fast and maintain poorly.

To look beyond the observed period, the team built an ensemble forecast combining three complementary statistical approaches: an ordinary least squares linear trend, exponential smoothing with a smoothing parameter of 0.7, and an autoregressive integrated moving average model of order (0,1,0) with drift. Each method captures different features of a time series. The linear trend extrapolates the average direction of change; exponential smoothing weights recent quarters more heavily, allowing the forecast to respond quickly to the post-2022 decline; and the ARIMA specification models the quarter-to-quarter random walk behavior of the counts. Averaging across such an ensemble hedges against the failure modes of any single model. The consensus projection is that quarterly crash numbers will likely remain high in the short term, though the authors emphasize that forecast uncertainty widens as the horizon extends, a reminder that statistical extrapolation is a guide, not a guarantee.

The authors are careful, and rightly so, to distinguish description from prescription. Their policy recommendations, a shift toward risk-based road safety strategies centered on speed regulation, vehicle safety checks and enhanced emergency response systems, are explicitly framed as inferences from the observed patterns rather than as evidence of intervention effectiveness. This is an important caveat. Knowing that northern states have high fatality rates per crash does not by itself prove that speed cameras or ambulance networks would reduce them; it identifies where such interventions should be trialed and evaluated. Nigeria’s road safety authorities now have a statistical map of where the lethality problem lives, but converting that map into saved lives requires controlled implementation and rigorous before-and-after measurement, the kind of evidence that has transformed road safety in countries like Sweden and Australia.

The study’s methodological choices also offer a template for other countries with fragmented road safety data. The fatality rate per crash is a deceptively simple metric with powerful diagnostic value: it separates the question of how often crashes happen from the question of how deadly they are when they do, and the two demand different remedies. Better road engineering, vehicle standards and post-crash care lower the fatality rate; traffic management, enforcement and driver education lower crash frequency. Nigeria’s flat fatality rate amid falling crash counts suggests that decades of effort have moved the first lever barely at all. The World Health Organization has long argued that post-crash response is the most neglected pillar of road safety in low- and middle-income countries, and this dataset gives that argument a precise national geography.

For a country of more than 220 million people, where roads carry the overwhelming majority of passenger and freight movement, the stakes of this research are hard to overstate. Nearly 6,500 deaths in a single year is not an abstract number; it is a rolling public health emergency that disproportionately strikes working-age adults, the demographic backbone of any economy. What this study delivers is clarity: a verified, quarter-by-quarter, state-by-state accounting of where crashes cluster, where they kill, and where the two diverge. The finding that lethality is concentrated in places with fewer crashes should redirect attention and resources toward the corridors and communities where a collision is most likely to end a life. The data now exist. The question the study leaves open, and one that Nigerian policymakers must now answer, is whether the will exists to act on them.

Subject of Research: Spatiotemporal analysis of road traffic crash fatality risk across Nigerian states from 2020 to 2024

Article Title: Spatiotemporal dynamics and fatality risk patterns of road traffic crashes in Nigeria (2020–2024)

Article References: Hassan, M., Nadarajah, S., & Ogunsola, I. (2026). Spatiotemporal dynamics and fatality risk patterns of road traffic crashes in Nigeria (2020–2024). Discover Sustainability. https://doi.org/10.1007/s43621-026-04760-y

Image Credits: AI Generated

DOI: 10.1007/s43621-026-04760-y

Keywords: road traffic crashes, Nigeria, fatality rate, spatiotemporal analysis, Federal Road Safety Corps, road safety, trend projection, ARIMA, exponential smoothing, public health, speed violations, vehicle safety