- Foundation
- About 14 min
- 5 practice questions
AI in farming and weather: crop photos, pest alerts and flood warnings
Describe how AI helps farmers detect crop disease from photos, get pest alerts, plan with soil and yield predictions, and how it supports weather and flood forecasting and drone use.
If something looks wrong, please tell us.
Farming is full of uncertainty: will it rain, will pests attack, which fertiliser does this field need, will the river flood? Farmers have always answered these questions from experience. AI adds another tool: it can study photos, weather records, soil tests and satellite images far faster than a person, and turn them into timely advice. Because agriculture supports a large part of India's people, this is one of the most important uses of AI in the country.
Kinds of use in agriculture
| Problem | How AI helps | What data it uses |
|---|---|---|
| Crop disease | Identifies a likely disease from a photo of a leaf taken on a phone | Thousands of labelled photos of healthy and diseased leaves |
| Pests | Warns when conditions favour a pest outbreak, or identifies insects caught in traps | Weather data, pest records, images of insects |
| Soil health | Suggests the type and amount of fertiliser for a field | Soil test results, crop type, past yields |
| Yield prediction | Estimates how much a crop will produce before harvest | Satellite images of crop health, weather, past yields |
| Irrigation | Advises when to water and how much | Soil moisture sensors and weather forecasts |
| Drones | Surveys fields from above and helps with targeted spraying | Aerial images analysed by AI to spot stressed patches |
How a crop-disease app works
How a crop-disease app works
- Photograph the leaf
- App compares with learned patterns
- Names the likely disease And how confident it is.
- Suggests what to do Often in the farmer's own language.
- Farmer confirms with an expert Before spending money on chemicals.
- The farmer photographs an affected leaf with a phone.
- The app compares the photo with patterns it learned from many labelled leaf images.
- It names the most likely disease and how confident it is.
- It suggests what to do, often in the farmer's own language, by text or voice.
- The farmer confirms with a local agriculture officer or expert before spending money on chemicals.
Seeing crops from space
Satellites and drones carry cameras that record light the eye cannot see. Healthy green plants reflect a lot of near-infrared light, while stressed or dry plants reflect less. AI models read these images over large areas and mark fields that are under stress, sometimes before the farmer notices anything from the ground. Combined with weather data and past harvests, the same images help estimate yields for a whole district.
Weather and flood forecasting
Weather forecasting traditionally uses physics-based computer models of the atmosphere. AI adds the ability to find patterns in huge records of past weather, satellite pictures and radar, which can sharpen short-range forecasts such as where heavy rain will fall in the next few hours. For floods, AI models combine rainfall forecasts, river levels measured upstream and the shape of the land to predict which areas may go under water and when. The value lies in early warning: hours or days of notice let people move, protect animals and save stored grain.
From data to a flood warning
- Data in Rainfall forecasts, river levels upstream and the shape of the land.
- AI flood model Predicts which areas may go under water and when.
- Early warning Hours or days of notice.
- People act Move, protect animals and save stored grain.
Worked example 1: spots on tomato leaves
Lakshmi, a tomato grower, sees brown spots spreading on leaves. She photographs a leaf in a crop-advisory app. The app reports a likely fungal disease with high confidence and suggests removing affected leaves and a suitable spray. Working: the app matched the pattern of spots with images it was trained on. Lakshmi checks with the local agriculture officer, who confirms the diagnosis. Why check? A similar-looking problem, such as a nutrient deficiency, needs a completely different remedy, and a wrong spray wastes money and harms the soil.
Worked example 2: a flood warning
Suppose heavy rain is forecast over a hilly catchment, and river gauges upstream show the water rising quickly. A flood model combines the forecast rainfall, the measured river levels and the slope of the valley. It predicts that low-lying villages downstream may flood in about a day and a half. Working: the district administration sends alerts by SMS and through local officials, and families move people, cattle and grain to higher ground before the water arrives. The AI did not stop the flood; it bought time.
Limits on the ground
AI advice is only useful if it reaches the farmer and fits the field. Poor mobile connectivity, advice that is not in the local language, very small and varied plots, and training data from different regions can all reduce accuracy. A wrong recommendation can cost a season's income, so AI works best as an adviser alongside experienced farmers and agriculture officers, not as a replacement.
In farming and weather, AI turns photos, sensor readings, satellite images and weather records into early warnings and advice. A person checks before acting.
Key takeaways: AI helps farmers identify crop disease from phone photos, anticipate pests, use fertiliser and water better, and estimate yields from satellite images. In weather and floods, it sharpens forecasts and gives early warning. Its advice depends on good local data, connectivity and the farmer's language.
Check your understanding
-
1. A farmer photographs a diseased leaf and an app names the likely disease. What did the app learn from?
- The farmer's bank statements
- Mandi price lists
- Many labelled photos of healthy and diseased leaves
- The farmer's voice recordings
Show answer
Answer: C. Crop-disease apps are trained on large numbers of leaf images labelled with the correct disease, so they can match a new photo to a pattern. Bank statements, prices and voice recordings do not show what a disease looks like.
-
2. Why do satellite images showing near-infrared light help monitor crops?
- Healthy plants reflect a lot of near-infrared light, while stressed plants reflect less
- Near-infrared light kills pests
- Near-infrared light makes it rain
- Only dry soil reflects near-infrared light
Show answer
Answer: A. Healthy green plants reflect strongly in near-infrared, and stressed or dry plants reflect less, so the images show crop health over large areas. Near-infrared light does not kill pests or cause rain, and the key signal comes from the plants.
-
3. What is the main benefit of an AI-based flood forecast?
- It stops the river from rising
- It decides the compensation each family gets
- It replaces the need for river gauges
- It gives early warning so people can move to safety in time
Show answer
Answer: D. A flood forecast predicts where and when water may rise, giving people time to move. It cannot stop a flood, it actually needs river-gauge readings as input, and it does not decide compensation.
-
4. Which data would an AI model most need to advise a farmer on the right amount of fertiliser?
- The farmer's mobile number
- Soil test results along with the crop type
- The colour of the tractor
- The distance to the nearest railway station
Show answer
Answer: B. Fertiliser advice depends on what nutrients the soil already has (soil test results) and what the crop needs. A phone number, tractor colour or railway distance tells nothing about soil nutrients.
-
5. An app suggests an expensive spray for spots on a crop. What is the sensible next step?
- Spray immediately, because AI is never wrong
- Ignore the app completely
- Confirm the diagnosis with a local agriculture officer or expert before spending
- Spray twice the suggested amount to be safe
Show answer
Answer: C. AI can confuse similar-looking problems, such as a disease and a nutrient deficiency, which need different remedies. A person should confirm before acting. Blind trust, ignoring the tool entirely or overdosing chemicals are all poor choices.
Sign in to keep track of the lessons you finish.