Syria 2026 Agricultural Shocks — Overview

Floods and crop fires against the 2025 record-drought baseline

Key findings at a glance

The 2026 floods and crop fires took back ~2.2% of Syria’s 2025 cereal floor — real, measurable, and heavily concentrated on the Euphrates, but a fraction of the recovery, not a second collapse.

  • ~25,925 t of cereal production lost (lower bound; range 3,471–39,946 t).
  • 2.16% of the 12-governorate 2025 drought baseline.
  • ~105,504 ha cropland flooded (3 Euphrates/Khabur AOIs) and ~10,533 ha burned (12 governorates).
  • Hardest hit: Raqqa lost 7.8% of its own baseline.

The spine: the 2025 baseline was a record drought; the 2026 shock is water (an anomaly on a desertification trajectory) plus crop fires. The question is how much of the tentative 2026 recovery the shock reversed.

Choropleth map of Syria's governorates shaded by 2026 cropland-production loss as a percent of the 2025 baseline. Raqqa and Deir ez-Zor on the Euphrates are darkest (about 7.6 to 7.8 percent), Hasakah is mid-toned (2.8 percent), and all other governorates are pale (under 1.5 percent). Latakia and Damascus City are grey, marked verified-excluded.
Figure 1: Where the 2026 loss falls: cereal-production loss as a share of each governorate’s 2025 baseline. The Euphrates corridor (Raqqa, Deir ez-Zor) carries the damage; Latakia and Damascus City are verified-excluded (grey). First-half-2026 lower bound.
Important

First-half-2026 snapshot — figures are lower bounds. Every headline here covers floods Mar–Jun and fires May 1 – Jun 12 2026. The summer harvest and fire peak (Jun–Aug) is unobserved, so all figures are lower bounds. Field and expert verification is the gold standard above remote-sensing self-consistency; the study is built to re-run post-harvest for the concluded full-year result (DEC-039).

The analytical spine

The 2026 flooding matters because it is an anomaly on a desertification trajectory, not a recurring hazard. The baseline is drought; the shock is water, plus crop fires. This study quantifies how much of Syria’s tentative 2026 cereal recovery the floods and fires reversed, measured as damaged cropland hectares translated into production and food-security impact against the 2025 drought baseline.

  • 2025 — record drought (baseline). FAO/GIEWS put the cereal harvest near 1.2 Mt, roughly 60%+ below average and the worst in about 60 years, with ~16.3M people at food-security risk.
  • 2026 — tentative recovery. Harvest projected to rebound toward ~2.3 Mt on more even rainfall, though sowing targets were not fully met.
  • 2026 — the shock. March and June Euphrates flooding submerged thousands of hectares of recovering cropland in Raqqa and Deir ez-Zor, and a national crop-fire signal (Hasakah-dominant) burned cropland across twelve governorates.

Headline — validated damaged cropland by governorate

Code 
import matplotlib.pyplot as plt
import seaborn as sns
from viz import apply_theme, caveat_footer, CAVEATS, PHENOMENON_COLORS, bar_value_labels

pal = {k.value: v for k, v in PHENOMENON_COLORS.items()}
apply_theme()
order = (impact.groupby("aoi_id")["damaged_ha_headline_union"].sum()
               .sort_values(ascending=False).index.tolist())
fig, ax = plt.subplots(figsize=(9, 5))
sns.barplot(
    impact, x="aoi_id", y="damaged_ha_headline_union", hue="phenomenon",
    hue_order=["flood", "fire"], order=order, palette=pal, ax=ax,
)
# Hatch + annotate the flagged Hasakah-June flood bar (the caveat should ride the
# figure, not only the prose — RQ1 flags it as a likely SAR harvest artifact).
flood_bars = ax.containers[0]  # hue_order[0] == "flood"
h_idx = order.index("hasakah")
bar = flood_bars[h_idx]
bar.set_hatch("////")
bar.set_edgecolor("white")
ax.annotate("flagged\n(likely artifact, RQ1)",
            xy=(bar.get_x() + bar.get_width() / 2, bar.get_height()),
            xytext=(0, 18), textcoords="offset points", ha="center", fontsize=7.5,
            color="#9b1c1c",
            arrowprops=dict(arrowstyle="-", color="#9b1c1c", lw=0.8))
bar_value_labels(ax, fmt="{:,.0f}", threshold=1000)
ax.set_xlabel("Governorate (area of interest)")
ax.set_ylabel("Damaged cropland — union headline (ha)")
ax.set_title("Validated 2026 cropland damage: floods dominate the Euphrates AOIs")
ax.tick_params(axis="x", rotation=35)
for lbl in ax.get_xticklabels():
    lbl.set_ha("right")
ax.legend(title="", loc="upper right", frameon=True)
caveat_footer(fig, CAVEATS["case_study_2026h1"])
plt.show()
Grouped bar chart of 2026 damaged cropland in hectares by governorate, split into flood (blue) and fire (red). Raqqa and Hasakah have the tallest flood bars (about 45,000 and 39,000 hectares), then Deir ez-Zor (about 22,000). Hasakah's flood bar is hatched to mark it as a flagged likely artifact. Fire bars are all small, with Hasakah's the largest at about 4,100 hectares.
Figure 2: Validated 2026 damaged cropland (union-headline ha) by governorate and phenomenon. Floods on the Euphrates dominate. The hatched Hasakah flood bar is flagged by RQ1 as a likely radar harvest artifact. First-half-2026 lower bound.

↓ Data behind this figure (impact_by_aoi.csv)

The loss is dominated by riverine flooding on the Euphrates: Raqqa is the single largest production loss, followed by Hasakah and Deir ez-Zor. One caution carried on the chart itself: Hasakah’s large flood figure is inflated by the June hectares that RQ1 flags as a likely radar harvest artifact, so it is the most uncertain of the three. Excluding those flagged hectares moves the study total only ~5% (see Food security), because the headline already counts each AOI’s peak single flood date rather than summing dates. Fires are spatially widespread (Hasakah-dominant) but each governorate’s burned cropland is small relative to the flood extent.

What the research questions found

RQ1 source split HIGH  ·  RQ1 natural-vs-managed release LOW  ·  RQ2 2026 conflict overlay DATA GAP  ·  RQ3 control comparison DESCRIPTIVE ONLY

  • RQ1 — Flood attribution. The June Euphrates flood is upstream-sourced (zero local rainfall, a sustained ~1,600 m³/s dry-season discharge, about 6× the drought baseline). Whether the release was natural or managed is LOW confidence: the reanalysis cannot resolve dam operations, the dry-season plateau is consistent with managed release, but no dam-release fraction is asserted.
  • RQ2 — Fire attribution. Live ACLED Syria coverage ends 2025-06-13, so the 2026 conflict overlay cannot be computed (a data gap, not a finding). The method, demonstrated on the 2025 window, shows cropland fires sit no closer to armed conflict than cropland in general, i.e. agricultural and seasonal rather than frontline-concentrated. There is no active armed conflict in the 2026 window.
  • RQ3 — Damage vs control areas. A descriptive overlay only (DEC-005). It maps where damage falls relative to indicative control zones and makes no differential or causal claim about either administration.

How to read this report

Every figure that makes a food-security or research-question claim is built only from human-validated records: the validation gate is enforced in code via viz.consumable_records(), and both pipelines had their gates closed by a human domain expert, never by an agent (DEC-007). Maps and attribution figures carry their caveats inline. Method codes like DEC-033 link to the plain-language glossary.

Show the validation gate applied to the raw damage tables 
from schema.damage_schema import read_csv
from viz import consumable_records

flood_recs = read_csv(OUT / "floods" / "flood_damage.csv")
fire_recs  = read_csv(OUT / "tables" / "fire_damage.csv")

for name, recs in (("flood", flood_recs), ("fire", fire_recs)):
    kept = consumable_records(recs)
    print(f"{name:5s}: {len(kept):>3d} validated / {len(recs):>3d} total records "
          f"reach the report")
flood:  63 validated /  63 total records reach the report
fire :   0 validated /  48 total records reach the report

Syria 2026 Agricultural Shocks · first-half-2026 snapshot (floods Mar–Jun; fires May 1 – Jun 12 2026) · figures are lower bounds (DEC-039). Author: Muhanad Abulhusn. Rendered from human-validated (Tier-2) records only.

Data sources: FAO/GIEWS (baseline & production), GloFAS (discharge), CHIRPS (rainfall), VIIRS/FIRMS (active fire), Sentinel-1/2 & Copernicus EMS (flood/burn extent), ACLED (conflict, cited not republished). License: CC BY 4.0.

Suggested citation:

Abulhusn, M. (2026). Syria 2026 Agricultural Shocks: floods and crop fires against the 2025 record-drought baseline (first-half-2026 snapshot). https://muhanad-husn.github.io/flood-fire/

Provenance & contracts: PRODUCT.md · STRUCTURE.md · decision log · source repo