The weather has a significant impact on the agriculture industry as it influences crop growth, water and fertilizer requirements, pest infestation, field activities and overall yield.  Farmers rely solely on expected rainfall and temperature levels during cultivation seasons to plan their field activities. Based on the experience of past weather conditions and their intensities, farmers decide when to plant and harvest, what crops to be grown, optimal soil structure, irrigation and fertilizer application. However, continuing climate change has caused drastic variations in weather patterns, making it increasingly difficult for farmers to predict the weather. Therefore, accurate and reliable machine-based weather forecasts are essential to adapt to adverse weather conditions and adjust the work plan and planting schedule accordingly. This weather information helps farmers to plan their activities efficiently in order to boost their yields and profits and minimize costs and losses.

Agriculture Industry in Sri Lanka

As stated by the Ministry of Agriculture, over 25% of the total population of Sri Lanka directly depends on agriculture. Hence, Agriculture Industry plays an important role in generating livelihood and employment opportunities, especially in rural areas. In 2020, agriculture accounted for 7.4% of the Gross Domestic Product of Sri Lanka (Ministry of Agriculture). However, being an island nation in the Indian Ocean, Sri Lanka is vulnerable to climate change and mostly affected by extreme weather events such as droughts and floods.

Recent trends of unusually high temperatures in April and May followed by over 300 millimetres of rain within 24 hours that induce catastrophic floods are concerning. As reported by Rajawat in her research, these weather swings directly affect the overall agricultural productivity and the economy of the country. According to the Sri Lanka National Report on Disaster Risk, Poverty and Human Development Relations, more than 95% of the agricultural crop losses in Sri Lanka are due to floods (52.2%), droughts (52.2%), and extreme wind events (4.2%).

Sri Lankan farmers are facing new challenges with the changes in weather patterns. Unpredictable monsoon seasons and rainfall intensities are causing some to abandon agriculture.  Bad weather can devastate the lives and economy of a country and it is one of the major driving forces behind poverty.

Issues with Current Weather Forecasts

The majority of existing forecasting models exclusively depend on conventional meteorological infrastructure and are not economical, sustainable and equitable. Due to the massive investment required to deploy and maintain traditional infrastructures such as weather sensors, satellites and radar systems, governmental meteorological sources have limited data points for forecasting.

This is common in low-income countries, where they have fewer live weather stations due to operating and maintenance costs. According to the World Meteorological Organization (WMO) and the World Bank, it requires an additional 4000 to 5000 basic observations across the African continent and it will need at least USD 1 billion to upgrade existing infrastructure and around USD 400 to 500 million per year for operations and maintenance.

As a result, the forecasting agencies are incapable of providing accurate forecasts and actionable insights to businesses and individuals. People have no confidence in existing weather forecasts for basic and crucial personal and business decisions in their everyday lives. This problem is exacerbated in emerging economies, where most people do not have access to accurate and credible weather forecasts and early warnings for floods and other severe weather conditions.

In addition, due to the lack of specialized meteorological equipment and timely and reliable weather data points, these forecast models provide low resolution forecasts (around 10 kilometre resolution), which are too wide to comprehend most weather phenomena. Hence, the traditional weather forecasting model is broken.

As a developing country, Sri Lanka faces similar problems with meteorological forecasts. In Sri Lanka, significant fluctuations of weather can be seen due to its topographical variations from the flat land to the central highlands. Within a few kilometres, you can witness drastic changes, from sunny weather to heavy rains. These fluctuations, which are common in mountainous areas are not subject to traditional low-resolution forecasts. Because of the low resolution and accuracy of existing weather forecasts, people believe that they are less reliable for basic personal and business decisions.

Low Resolution (13 km) model vs High Resolution model (3 km). Source: IBM GRAF

Weather of Things

Hyper-local but timely and reliable weather data gives a closer understanding of weather conditions in a particular location and assists in delivering accurate high-resolution weather forecasts. Advanced technologies such as Internet of Things (IoT) and machine learning (ML) provide a new opportunity to capture weather data from a variety of sources at a micro level.

This new approach, called “Weather of Things (WoT),” is all about capturing data from a wide range of modern infrastructures such as airplane sensors, pressure sensors from mobile devices, weather balloons, cameras, other IoT devices and cell tower transmissions along with traditional infrastructures such as satellites, ground radar, weather stations and translating them into valuable weather observations. Hence, it allows us to use various other data sources as weather sensors and superimpose more granular data into weather forecasting models.

Weather of Things – Network of Weather data

Opportunities for the Plantation Sector

WoT along with cutting-edge machine learning can enhance weather forecasts to provide localized actionable insights for the plantation sector. These insights allow planters to make better operational and business decisions.

• Field Operations – Determine the suitable time for fieldwork and plan day-to-day activities more efficiently.
• Irrigation – Decide when, where and how much to irrigate.
• Fertilization – Determine appropriate fertilizer form, application rate and application time to avoid wastage of resources and money.
• Pest and Diseases – Predict and determine when the chemicals should be applied to control the growth of pests and diseases.
• Crop Growth – Understand and track the growth status of crops.

The plantation sector in Sri Lanka can benefit from a local weather platform that incorporates multiple sources of weather data to generate granular and accurate weather forecasts. Aforesaid actionable weather insights will empower “precision agriculture”, a management strategy that focuses on the accuracy and control of crop production. In fact, with more productivity gains and fewer crop damages, it will boost profits and improve the livelihoods of farmers. Hence, it plays a significant role in the evolution of the agriculture industry. Besides, granular and accurate forecasts could also benefit aviation, entertainment, utilities, sports, logistics and other industries.

 – Written by Inuri Muthukumarana.