In a not too distant future, in a galaxy very similar to ours, the statistics will be all remote. Satellites will take photos and scan the entire surface of a given country, or a region. Algorithms tuned for many years with intense validation work in the field, will prepare maps of agricultural areas, livestock, forest, fallow land and other types of uses. Based on these maps which can be compared instantaneously with maps of past years, the sampling frame will be determined to design a survey, be it of production, estimation of crops, yields, and even of losses due to a pest or special event (for example, a frost).
But that’s not all, infrared images of those same areas will help statisticians to prepare maps of each crop, based on the wavelength that each species reflects and, with some more work, of each variety. Depending on the month or the day the photo is taken it will be possible to know if the crop has a week or six weeks. You can estimate average performance instantly, thanks to those wonderful algorithms that calculate everything in less time than it takes to take a breath.
With these crop maps you can calculate specific samples by crop which complement the master sample frame, and in the same way as in maps of agricultural areas, you can instantly compare with maps of past years to analyze changes in time and further refine the accuracy of the estimates. Farmers will be contacted directly by the software to complete the information, which will be sent by mobile devices to a centralized database. And the satellite images used can be complemented with photos and information from drones and remote sensors. It will be possible to estimate the water consumption, emission or mitigation of greenhouse gases, changes in the maturation and flowering of species or grazing patterns.
Both maps of agricultural areas (macro) and crop maps are calibrated and validated in the field through an intense manual work. But little by little that information is accumulated, the coefficients of a region turn out to be are very similar to those of other regions… and like the rankings of the search engines of the Internet, tend to the same number in the long term. After a few years such intense validation is no longer necessary, the confidence in the algorithm in the satellite sensor will be so high that the results of this analysis will suffice. It is the closest thing to artificial intelligence that can be for agricultural statistics.
Is it the future or is it the present? All that technology is at hand. Satellite and radar images are available to the public at no cost. Algorithms can be downloaded without any cost. What is not yet available is the total validation in the field. But that future, which is the dream of all those who work with statistical data from the agricultural sector, is already at hand.
Some of these initiatives are being implemented through the Project for the Improvement of the Agricultural Statistical Information System and the Agricultural Information Service for Development of Peru (PIADER), which is being implemented by the Government of Peru and co-financed by the Inter-American Development Bank (IDB). The Project finances the preparation of the statistical sampling frame and the agrarian national surveys and the strengthening of the public and private capacity to offer agricultural statistical information of the highest quality. Even if we attach a RFID chip (Radio Frequency Identification Chips) or a GPS to each animal that grazes in the paddocks, we can do the same for the livestock sector. And although it looks like science fiction, this type of grazing optimization is being implemented through a BID Lab project that is being executed in Peru, in collaboration with a cooperative of wool producers of Alpaca in Arequipa.
We are not far from stop using paper and field surveys to generate the information needed by public and private sector decision makers. And this aggregation of information will allow countries to define more precisely where investment is needed and what kind of policies should be implemented to achieve greater growth in the agricultural sector. More information and more accuracy translate into better and greater investments in the sector, which in turn render higher productivity and better income for producers. And if we also consider the savings in paper and fuel for travel, maybe these initiatives would also be environmentally sustainable!