A rainfall deficit describes how much rain is missing relative to a reference period. It does not, by itself, tell you how much water remains available to a farm or community. Reading drought reports well means following water beyond the rain gauge.
This explainer separates rainfall, soil moisture and water-system indicators, then shows how to compare them. It is not a current drought alert for the Mekong basin, Bali or any other location. The featured photograph is a historical illustration of rice farming, not evidence of drought conditions.
Three drought questions that need different measurements
The US National Integrated Drought Information System distinguishes several types of drought. Meteorological drought concerns dry weather; hydrological drought concerns shortages evident in the water system; agricultural drought concerns effects on crops. These descriptions help explain why two reports about the same region can emphasise different conditions.
Start by identifying the question a report actually answers:
- Rainfall: How unusual was precipitation over the stated area and period?
- Soil moisture: How much water is present at the measured or modelled depth?
- Water supply: What do river, reservoir or groundwater records show about the system being assessed?
Do not promote an answer to one of those questions into an answer to all three. A rainfall map is useful evidence about rainfall even when it cannot establish the condition of a particular irrigation system.
Check what the rainfall percentage is compared with
A percentage becomes meaningful only after its reference is stated. Check the geographical area, observation dates and baseline before comparing two rainfall headlines.
Hypothetical example: Suppose a district receives 120 millimetres of rain during a month for which its chosen long-term average is 200 millimetres. The observed amount is 60% of that average, and the deficit is 40%. These are two descriptions of the same calculation:
- Share of average: 120 ÷ 200 × 100 = 60%.
- Deficit relative to average: (200 − 120) ÷ 200 × 100 = 40%.
Neither figure means that reservoirs are 40% empty or that crop yields fell by 40%. Those outcomes were not measured in this example. Nor does the calculation alone establish a drought category: that requires the classification method and relevant context.
The figures are invented for explanation. They are not observations from a real district or a forecast for the next season.
Soil moisture adds information, but scale matters
NIDIS explains that soil moisture depends on soil and vegetation as well as weather. Water can enter the soil, move deeper or across the surface, be taken up by plants, and return to the atmosphere through evaporation and transpiration. A rainfall total therefore does not describe every part of that movement.
The measurement method also matters. An in-ground sensor represents a small volume of soil at a particular depth. Models and satellite estimates cover broader areas with their own limitations; many remote-sensing methods describe only the uppermost soil layer. Before applying a regional map to a field, check its depth, resolution and validation information.
A practical reading habit is to write those details beside the reported value. “Soil moisture is low” leaves more uncertainty than a statement that identifies the layer, location, date and comparison used.
Why one wet spell may not settle the question
NIDIS guidance on drought timescales notes that wet spells can occur within a drought. It also explains that groundwater can respond more slowly than surface conditions, while pumping can affect the amount available. Different indicators need not recover together.
When a headline says rain has ended a drought, ask which indicator changed and over what period. Has the report documented rainfall recovery, improved soil moisture, replenished storage, restored supply, or several of these? That question helps clarify the claim without assuming that the rain was unhelpful.
A Mekong example: use rainfall and river information together
The Mekong River Commission’s forecasting programme combines river monitoring with flood and drought forecasting services. It provides rainfall and water-level information, historical comparisons and forecast bulletins. Its explanation of mainstream monitoring distinguishes stations serving different parts of the basin, including Chiang Saen, Pakse and Kratie.
For readers following the Mekong, this gives a useful route from a broad headline to more specific evidence. Identify the station or area relevant to the claim, check the observation date, and separate a recorded value from a forecast. A chart for one station should not silently become a measurement for the whole basin or for every farm nearby.
This is an explanation of the monitoring approach. No live station reading has been used here to classify today’s conditions.
A short checklist for the next drought story
- Variable: Is the story measuring rainfall, soil moisture, river flow, storage or an observed impact?
- Place: Is it a point measurement, a district estimate or a basin-wide summary?
- Period: Does the number cover a day, month, season or longer interval?
- Comparison: Which reference period or classification threshold is used?
- Evidence type: Is the value observed, modelled or forecast?
- Limit: What remains unknown about access to water or effects on users?
Our guide to alternate wetting and drying in rice farming explores why control over irrigation matters when evaluating a water-management practice. For a broader introduction to checking measurements and reference periods, see the climate evidence guide.
The most useful drought explanation keeps the weather signal, the condition of the water system and the reported impacts distinct. Connecting them requires evidence; treating them as interchangeable can obscure the problem people actually face.