Localized Climate Forecasting System Seasonal Climate and Weather Prediction for Farm Level Decision Making

R. Rengalakshmi



Recent developments in weather and seasonal rainfall prediction have increased the accuracy and reliability of forecasts of the Indian monsoon. Despite these advances, availability and access to location-specific forecasts to take proper decisions at the farm level is very limited. Traditionally farmers in India have been using a set of indicators that have varied levels of dependability for rainfall prediction and have evolved several coping strategies and mechanisms.

The M. S. Swaminathan Research Foundation (MSSRF), based at Chennai, India initiated a project on "Establishing decentralized climate forecasting system at the village level" to create and enhance farmer's capacity to use locale-specific seasonal rainfall and weather forecasting in collaboration with Reddiyarchatram Seed Growers Association (RSGA), a farmers association at Kannivadi in Dindigul district of Tamil Nadu state, India. The main goal of the project is to create an access and enhance farmers' capacity to use location specific seasonal climate and weather predictions to improve their livelihoods. The major objectives are to study the seasonal climate variations and chronicle the farmer's traditional coping strategies and knowledge. The study also aims at evolving a methodology for downscaling with appropriate institutional linkages and converting the generic data into location-specific, medium term, inter- and intra-seasonal climate and weather forecasts. Probabilistic seasonal climate and weather forecast information is translated into appropriate farmer friendly versions for its practical use in crop management.


Study Area

Reddiyarchatram block is a semi-arid region located in Dindigul district of Tamil Nadu, India, covering a geographical area of 280 km2. More than 80% of the households in the district depend on agriculture. Important planting seasons are June-July and October-November for both the irrigated and rainfed crops, in addition to the summer irrigated crop. The mean annual rainfall is 845.6 mm. Rainfall in the region is characterized by a large variation between seasons. Though the area benefits both from the northeast monsoon (October-December) and the southwest monsoon (June-September), maximum percentage (52.5%) of rainfall is received during the northeast monsoon and nearly 25.8% of the total annual rainfall is received during the southwest monsoon. The area receives only 5.4% of the total annual rainfall during January and February and nearly 16.3% during the summer seasons between March and May. The total area under cultivation is 24 624 ha which includes both dry and irrigated lands. Approximately 29 600 households are involved in agriculture and more than 50% of the households are small and marginal farmers. Sorghum, small millets, grain legumes, cotton and chickpea are the major annual crops cultivated under rainfed conditions. Cotton, maize, flower crops, vegetables, gherkins, sugarcane, annual moringa, paddy, onion, etc. are the most important annual crops grown in this region. The major source of irrigation is underground water through wells followed by small tanks and reservoirs.



The study was initiated during October 2002 to March 2004 in five villages where Village Knowledge Centers (VKCs) are functioning. The computer based Village Knowledge Centers with Internet connection provides static information about the agronomical practices of the different crops cultivated in the region and the dynamic information like price details of the main agricultural produce from different markets, availability of inputs, farmers entitlements, etc. A set of VKCs are operating in the region connected with a 'hub' in the center and the 'hub' is the nodal point, which receives the generic information and adds value by converting it to local specific information. The local community manages the VKCs; access is ensured to all irrespective of caste, class, gender and age. Need based content creation is being regularly done on the basis of the feed back from the local women and men farmers. The local village people have been trained in the management of modern information and communication technologies including networking.

In each village, traditional knowledge system on weather and climate forecast was studied through conventional survey using questionnaire, anthropological tools such as participant observation, and participatory developmental tools such as Venn diagram and Focus Group Discussions (FGD). The traditional weather and seasonal rainfall predictors were studied among the selected sample households through questionnaires. Anthropological tools such as open-ended interviews were used to study the metaphors, folklore and proverbs that gave a better perspective on the traditional knowledge. A series of Participatory Rural Appraisals (PRAs) were organized in representative villages in the block that focused on the social system, existing natural resources, agricultural seasons and rainfall patterns and also on the prevailing pattern and system of information flow. The needs, constraints and coping strategies on weather and climate of farmers and agricultural laborers were assessed through FGD and these views were triangulated through informal discussion with knowledgeable men and women farmers.

MSSRF facilitated linkages to get the scientific forecast between hub of the VKCs and National Centre for Medium Range Weather Forecast (NCMRWF) for medium range weather forecast and the Tamil Nadu Agricultural University (TNAU) for seasonal rainfall forecast. The hub center manages a 'B' observatory; animators were trained in observatory management with the technical support of TNAU. They regularly record the local weather parameters (maximum and minimum temperature, soil temperature at different depths, sunshine hours, wind direction and velocity, evaporation rate, relative humidity) according to the norms of Indian Meteorological Department in the prescribed format and communicate the same to NCMRWF twice a week through electronic mail. In turn NCMRWF provide weather forecast twice a week to the hub center on cloud cover, precipitation, temperature, wind direction and wind velocity. Similarly, linkages were established to receive the seasonal rainfall forecast from TNAU.

The hub center receives the forecast and converts the generic information received from these two institutions into location-specific farmer friendly language (for example if the wind direction is 100°, it is communicated to the particular village in their local parlance) and disseminates the information to farmers and agricultural laborers through VKC, bulletin boards and local newspaper to the farmers.

Initially MSSRF trained the animators to convert the generic information into farmer friendly versions. The information is being communicated to other VKCs through fax mode and can be accessible through multimedia folders using Internet. The messages are communicated to nearby villages by the VKCs through bulletin boards that are located in 15 different villages. A Focus Group Discussion was carried out in each of the villages with the men and women to communicate the forecast. Initially we explained the method by which the scientific forecasts were generated and its attributes to the farmers. The probabilistic nature of the seasonal rain forecast was explained to the farmers, and simple locally familiar games were organized to clearly explain the concept of probability. Then using the climatological data analysis 'probability of exceedance' graph was generated to explain the relationship between rainfall amount (forecast) and probability. Attempts are being made only to communicate the forecast information to the people instead of giving follow-up advisories. It allows the farmer to take their own decisions, because under the varied cropping pattern and rainfed situations, farmers take decisions based on the event of rainfall and follow dynamic strategies instead of a single strategy as most of the forecasters recommend. The entire process is institutionalized through these VKCs.


Results and Discussion

Understanding people's perceptions and knowledge of weather and climate is critical for effective communication of scientific forecasts. The knowledge is learned and identified by farmers within a cultural context and the knowledge base follows a specific language, belief and process. The local men and women members assess, predict and interpret by locally observed variables and experiences using combinations of plants, animals, insects, and meteorological and astronomical indicators. Farmers use different kinds of traditional knowledge for rainfall prediction based on their observation with different types of phenomena like wind movement, lightening, animal behaviors, birds movement, halos/rings around the moon and the shape and position of the moon on 3rd to 5th day from the formation, etc. This type of knowledge provides a framework to explain the relationships between particular events in the climate and farming. Farmers use different types of predictors (based on environmental and biological criteria) in combination to take critical farming decisions and to decide on adaptive measures. The knowledge is evolved by locally defined conditions and needs, in other words this knowledge is context specific.

Men and women have different kinds of knowledge and use it for different purposes. Similarly village elders are more knowledgeable and are able to use more indi cators with greater understanding of the reliability of various indicators. The older men and women were able to provide more than 12 indicators with different lead times, whereas the middle aged (25 to 35 years) persons could provide only 3-4 indicators. Farmers as well as agricultural laborers have their own indicators that are based on their need and interaction. Also, farmers are able to provide more indicators than the agricultural laborers. The variations in the indigenous knowledge in a community are based on age, gender, caste, class and literacy.

The indicators clearly show that this indigenous knowledge on seasonal rainfall and weather is qualitative in nature. Weather predictions are used to take short-term decisions both in the irrigated and rainfed systems. It helps the small and marginal farmers to plan various agronomic practices more effectively especially at the time of sowing, weeding, spraying of chemicals and harvesting and post harvest operations. However, farmers use seasonal rainfall predictions to prepare themselves for anomalies related to rainfall. For example it helps to decide the cropping pattern for that season, if the rainfall is normal, they can go for high value crops like maize with high yielding varieties, otherwise if it is below normal they can plan for short duration drought resistant pulses and small millets. Farmers have been using different strategies to adapt and cope up with uncertain weather and climate based on their experience and acquired knowledge from previous generation. The important decisions are selection of cropping system, mobilizing seed, fertilizer and application, decisions on sowing (early or late), land and bed preparations, mid season corrections such as reducing population/providing irrigation. Similar to the seasonal forecast, weather forecast is being useful for the small and marginal farmers to plan the agronomic practices more effectively especially at the time of sowing, weeding, spraying of chemicals and harvesting and post harvest operations.

In the Focus Group Discussion farmers expressed that the increasing variability in rainfall have reduced the farmers' confidence in their own predictors and hence they are increasingly looking for scientific forecasts. They expressed the variability in terms of more water deficit years, late onset of rains and premature end of rains, and irregular distribution in time and space. Climatological analysis of the inter annual variability using 20 years of annual rainfall in this region indicated that the variability was about 36% and across the seasons the variability in terms of CV is high during the southwest monsoon season (71.6%) followed by the northeast monsoon season (52.2%). Hence, the challenge and necessity is to provide reliable forecasts through appropriate methods based on the needs of the farmers.

During 2003 and 2004 winters, monsoon rainfall amount was predicted and communicated to the farmers. Based on the two years experience, farmers indicated that it is very difficult to take decisions in the farm based on this forecast information. Instead, it might help them to prepare against anomalies in the future, provided the forecasts are accurate over years. Though farmers are listening and carefully monitoring the correlations, they expressed that they need time to observe the effectiveness of scientific forecasts over seasons or years. Based on the request of the farmers, four rainfall measuring devices were installed in different villages in this region and the rainfall was carefully recorded by the Knowledge centers.

Farmers expressed that their traditional practice follows dynamic strategies based on the event of rainfall, which is completely different from following a single strategy based on one prediction before the crop cultivation. They expressed that their existing strategies are more practical, evolved locally over years through trial and error considering the available natural resources. Thus the forecasts of a single rainfall amount do not support taking any short-term (e.g. like crop variety or plant population per unit area) or long-term decisions (like cropping system: monocropping or mixed cropping, etc.). Another important issue is that the probabilistic mode of the total amount of rainfall does not support farmers' need in terms of time of onset of rainfall and its distribution. It is one of the significant variables requested by the farmers to make decisions on initial agricultural activities, which may help to reduce the risk. Though farmers could understand the probabilistic nature of the rainfall over season, they expressed that it is very difficult to operationalize it, since it is not providing confidence (moral support) to the farmers, instead it indicates the lack of certainty and based on this they could not take major decisions. Also the two years experience indicates that learning takes time (observation over time/seasons) and the use has to do with familiarity.

With regard to the medium range weather forecasts, attempts are being made only to communicate the forecast to the people instead of giving follow-up advisories based on the forecasts. It allows the farmer to take decisions based on his/her field conditions. This is because under local situations, due to the heterogeneous nature of the field and crop conditions farmers take decisions based on the event and they have been following dynamic strategies instead of a single strategy which the forecasters recommend. A survey was organized to know the impact of the forecast information and nearly 66% of the farmers expressed that they have used it for taking farm management decisions. Around 72% of the farmers expressed the need for receiving forecasts at a much longer lead time interval, mostly 10 to 15 days.


Preliminary Conclusions

The study clearly brought out the importance of the vast traditional knowledge of the farmers on rainfall prediction and their understanding of its reliability through their observation, experience and practice in the field. The social stratification influences the evolution and management of knowledge. Understanding the local people's perceptions on rainfall prediction is necessary to communicate the scientific forecasts, since it is learned and identified by farmers within a cultural context and the knowledge base follows the specific language, belief and process. Intensive participatory dialogue between the scientific knowledge providers and user group's helps to define the strategies for using the forecasts in combination with traditional knowledge and skills. The project helped us to understand that, to develop a decentralized forecasting system at the village level needs a participatory approach to mobilize the farmers around the technology. On the other hand, access, availability of infrastructure, skill and expertise are crucial to develop reliable region-specific scientific forecasts to serve the farming societies. Farmers may not heavily rely on scientific forecasts until the forecasts have proven its reliability. At this phase due to the limited experience and observation it is difficult to derive any conclusion. It helps us to set the system and in the process slowly build up the farmers' understanding and confidence in scientific forecasts.


The author would like to thank Dr James Hansen, IRI for Climate Prediction, New York, USA and Dr Roland Fuchs, Director, START Secretariat, Washington DC, USA for providing the opportunity to undergo training and subsequently to carry out the project. The financial support provided by the David and Lucile Packard foundation, USA is gratefully acknowledged. The author extends her sincere thanks to Dr. Sulochuna Gadgil, IIS, Bangalore for providing technical and moral support and fruitful mentor-ship for the project. Author expresses deep gratitude for the technical support extended by TNAU, NCMRWF and Indian Institute of Tropical Meteorology (IITM) and the local farmers' association for actively taking part in establishing a decentralized village level forecasting system. The author sincerely thank Prof. M. S. Swaminathan, Chairman and Dr. K. Balasubramanian, Program Director of M. S. Swaminathan Research Foundation for their encouragement and guidance on conceptualization and implementation of the initiative.

Was this article helpful?

0 0
Renewable Energy Eco Friendly

Renewable Energy Eco Friendly

Renewable energy is energy that is generated from sunlight, rain, tides, geothermal heat and wind. These sources are naturally and constantly replenished, which is why they are deemed as renewable.

Get My Free Ebook

Post a comment