New Risk Index Tracks Jordan’s Food Security Year by Year

Jordan sits at one of the sharpest intersections of water scarcity, climate exposure, and food-import dependence in the world, and a new open-access study argues that the country needs a way to see agricultural risk coming before it becomes a crisis. In the Journal of Agriculture and Food Research, Huthaifa Alqaralleh and Audai Al-Majali introduce AgriShield-JO, a composite agricultural food-security risk index built annually for 2007 through 2024. Rather than producing a single opaque number, the framework is designed to tell policymakers not only how risky a given year was, but which mechanism drove the risk: climate and water stress, unstable domestic production, food-price pressure, or a gap in the system’s capacity to absorb shocks.

The starting point of the study is a conceptual one. Food security is conventionally described through four dimensions—availability, access, utilization, and stability—and the authors argue that stability has become the hardest to secure because modern food systems face simultaneous climate, market, geopolitical, and resource shocks. In a water-constrained economy like Jordan, a good rainfall year can still be undermined by price spikes, and stable prices can coexist with deepening long-term vulnerability from irrigation shortfalls. A single indicator, the authors contend, cannot capture both the level and the source of that risk. Their answer is a four-pillar architecture in which each pillar is scored on a 0-to-100 scale and then combined into a headline index, with the pillar scores preserved for diagnosis.

The technical construction is deliberately transparent. The climate-water pillar combines rainfall deficit and heat anomalies, drawn from NASA POWER observations at five representative locations—Aqaba, Karak, Amman, Irbid, and Mafraq—and evaluated against a 2000–2024 climate baseline. The production pillar converts yields for seven crops with consistent histories, including wheat, barley, tomatoes, cucumbers, potatoes, olives, and grapes, into a normalized crop-yield index. Market-price pressure is proxied by annual food CPI inflation from Jordan’s Department of Statistics, and the adaptive-capacity gap is measured by the share of cultivated land that is not irrigated. All indicators are direction-adjusted so that higher values mean higher risk, normalized with min–max scaling, and aggregated with theory-informed weights of 30 percent each for climate-water stress and production vulnerability and 20 percent each for market pressure and adaptive capacity.

The empirical results trace a strikingly non-linear risk trajectory. The index identifies 2008 and 2010 as the two highest-risk years in the sample, with scores of 69.4 and 60.1 respectively—both in the high relative-risk band. The 2008 peak coincides with the global food-price crisis, and the pillar decomposition shows that production and market pressures jointly produced that episode, with both pillars reaching their sample maxima that year. In 2010, by contrast, climate-water stress becomes the dominant driver, consistent with contemporaneous evidence that drought in the Jordan Valley constrained cultivated area. After 2010 the index declines gradually, with most years from 2011 to 2018 in the moderate range and the lowest scores in 2019 and 2020, before rising again to 36.2 in 2024, driven increasingly by climate-water and adaptive-capacity contributions.

Perhaps the most policy-relevant finding concerns which pillar dominates most often. Adaptive-capacity risk, measured through the irrigation gap, is the most frequent dominant driver, appearing in eight of the eighteen years—44.4 percent of the sample. Production risk dominates in five years, climate-water risk in three, and market-price risk in two. The authors are careful to explain the mechanism: a persistent irrigation gap can dominate in moderate-risk years when other pressures are low, whereas extreme production and market episodes generate the largest aggregate peaks. This distinction matters because the appropriate policy response differs fundamentally depending on the driver—yield stabilization for production risk, price surveillance and affordability support for market risk, drought preparedness for climate risk, and irrigation efficiency and resilience investment for adaptive-capacity gaps.

Beneath the national index, the study adds a crop-level vulnerability ranking that reveals how unevenly risk is distributed across Jordan’s agriculture. Combining yield volatility, production volatility, the frequency of negative yield growth, and average yield risk, the ranking places grapes first, followed by wheat, barley, and olives. Wheat and barley are particularly consequential because they are rainfall-sensitive staples, while grapes and olives are perennial tree crops whose exposure to accumulated heat and water stress plays out over longer adjustment horizons. By contrast, tomatoes, potatoes, and cucumbers record the lowest composite vulnerability scores, indicating greater historical stability under the selected metrics. The ranking survives a direct sensitivity test: replacing the baseline component weights with equal weights leaves the ordering essentially unchanged, with a Spearman correlation of exactly 1.000.

A substantial portion of the paper is devoted to the question that plagues all composite indicators: are the results an artifact of methodological choices? The authors run an unusually thorough battery of checks. Equal pillar weights produce an index nearly identical to the baseline, with a Pearson correlation of 0.993, and 2008 and 2010 remain the two highest-risk years. An entropy-weighted alternative, which shifts the weights dramatically to 57.1 percent for climate, still preserves the two peak years. A z-score normalization transformed through the normal cumulative distribution yields a correlation of 0.974 with the baseline. Leave-one-indicator-out tests, in which each component is removed one at a time, produce correlations between 0.825 and 0.976 without ever displacing the two principal stress episodes.

The Monte Carlo analysis quantifies weight uncertainty directly. In 100,000 draws, each baseline pillar weight is independently perturbed by plus or minus 25 percent and the weights renormalized. The 2008 score retains its risk category in 99.55 percent of draws, and 2010 in 100 percent; the only material instability appears in 2019 and 2023, whose baseline scores sit close to the 25-point boundary between the low and moderate categories. This pattern shows that classification uncertainty concentrates at category boundaries rather than in the identity of the main stress episodes. A scenario exercise for 2024, while explicitly a mechanical stress test rather than a forecast, illustrates the framework’s diagnostic use: a combined drought, heat, and market shock would push the 2024 score from 36.2 to 45.2, while a climate-smart investment scenario would lower it to 32.2.

The authors are candid about the limits of what they have built. The index covers only eighteen annual observations, too few for formal structural-break testing, statistical inference, or predictive validation. Min–max scaling is sample-dependent, so the 0-to-100 categories represent relative severity within the 2007–2024 calibration period rather than externally validated famine or crisis thresholds. Adaptive capacity is captured only through the irrigation-gap proxy and does not measure institutional quality, finance, extension services, or governance. Household access, nutritional utilization, and trade dependence—an omission the authors acknowledge as significant for a water-scarce, import-reliant economy—remain outside the balanced baseline because consistent annual series are unavailable. The 2008 and 2010 peaks are validated only retrospectively, against documented food-price and drought evidence, which the authors describe as face validity rather than forecast skill.

Even with those caveats, AgriShield-JO offers a replicable template for a class of countries that global indices struggle to serve. Broad benchmarking tools such as the Global Food Security Index, the FAO’s Suite of Food Security Indicators, and the INFORM humanitarian-risk framework are valuable, but they are not designed to explain the year-to-year agricultural mechanisms operating inside a single water-scarce economy. By decomposing risk into its constituent pressures, linking the national profile to crop-specific vulnerability, and making methodological uncertainty explicit, the framework functions as a monitoring dashboard rather than a stand-alone alarm: observe the aggregate score, identify the dominant pillar, inspect the underlying indicator, and match the intervention to the mechanism. For Jordan, where the National Food Security Strategy 2021–2030 has made resilience a strategic priority, and for other dryland economies facing similar compound pressures, that diagnostic clarity may prove as valuable as the numbers themselves.

Subject of Research: A composite agricultural food-security risk index for water-scarce Jordan

Article Title: AgriShield-JO: A composite agricultural food-security risk index for Jordan

Article References: Alqaralleh, H., & Al-Majali, A. (2026). AgriShield-JO: A composite agricultural food-security risk index for Jordan. Journal of Agriculture and Food Research, 31, Article 103327. https://doi.org/10.1016/j.jafr.2026.103327

Image Credits: AI Generated

DOI: 10.1016/j.jafr.2026.103327

Keywords: food security, Jordan, composite index, water scarcity, climate risk, crop vulnerability, irrigation, food prices, adaptive capacity, risk monitoring, agriculture, drought