Arid Regions Problems Can be Solved with the Help of Remote Sensing and AI

by | Sep 22, 2017

In recent years many areas of the US territory suffer from a drought of varying intensity. It is difficult to forget the consequences of the Great Drought of 2015, which was considered the largest in the last 1200 years.

As of July 2, 2017, according to the US Department of Agriculture, the abnormal dryness and some amount of moderate drought (D1) expanded in parts of the Northern Rockies and Northwestern U.S., with short-term drought impacts being more prevalent than long-term drought impacts. Several large wildfires developed in these regions, concomitant with the encroachment of abnormally dry and moderate drought conditions. On average, drought conditions worsened slightly in the Southern Plains, with parts of central Kansas, central Oklahoma, and South Texas showing deterioriation. Drought improvement was observed in a fairly limited area of the Southwest, stretching from southern portions of Arizona and New Mexico into west Texas.


This situation affects not only agriculture but also the quality of life in the cities. In such conditions, the issue of water saving becomes a sticking point and causes many problems in densely populated areas of the country. About 9 percent of the contiguous US fell in the moderate to extreme drought categories (based on the Palmer Drought Index) at the end of July ( NCEI Drought Report July 2017).


Is it possible to reach a peace agreement with the population?  

If it isn’t too subversive to say so, the state authorities are aware of the fact that loyal methods and propaganda do not bring proper results, not mentioning the fact that the mass installation of water meters can not be influenced. Water Districts tried to recalculate the indicators and bill residents post-factum, adding the cost of overruns to the general account. This leads to protests – no one wants to pay more!

According to Prop 218, Water Districts are not profitable organizations. The change in tariffs requires justification, for this purpose, it is necessary to provide a new method for planning the flow of water, and calculating wasteful use.

How to recognize the perpetrator?


One of such approach is to calculate the average flow of water resources per district and to determine the extravagance of consumers by identifying “more than twice the average.”

We can calculate water consumption in the regions accurately, using only the latest technologies and remote sensing data (RS), meteorological data and data from meters for estimating and planning the flow of water resources.

The human factor not only affects the accuracy of the data but also causes distrust of the population and the court proceeding afterward. For example, a human can not always recognize where the pool is, where the plants are, and where the tracks are using the pictures of the territory. In addition, the water management inspector can not obtain this information, without violating the sovereign right of private property (without intrusion into private property for measurements). This approach is extremely inaccurate, and not fair in relation to consumers who have a high flow rate due to the evaporation of water from a large area of green plantations.  

The newest technology was developed by California-based company EOS. WaterCloud service in cooperation with Google Maps and California’s Water Districts with the help of the best quality images classifies objects in the private sector and separates the information on the areas of foliage of plants and pools from roofs, paths and buildings and compares it with data on evaporation or use of water to care for a certain type of plant. Then, the planned flow rate is compared with the counters, which allows identifying wasteful consumers and marking their areas on the district map, for example, in red.


The heavy artillery for data processing – algorithms and artificial intelligence

It’s hard to believe, but such arrays of graphic and statistical information can be processed, calculate and analyze the total area of each kind of surface on a scale of the whole district.

The calculation of water consumption uses the methodology approved by Department of Water Resources. For indoor use, the daily water consumption per capita is 55 gallons. For calculation WaterCloud uses the data about the requirements for the incoming file with the initial data from the water district: APN – subscriber number of the site / unique parcel ID; postal address, street, city state, ZIP-code; registered number of residents; data from the meter in gallons; the size of the plot Ft2, etc. To assess outdoor use, the map of the imported data to the sites is made through APN; the total water consumption in the building, etc.  


Who can become a pioneer of the technology?

EOS company using WaterCloud technology is ready to solve a number of problems with water overflow in Water Districts of California (Santa Rosa, Marine Water, and East Bay) even today. A comparative assessment of the accuracy of automatic classification of buildings, vegetation, and water objects using multispectral aerial images and neural network. The average value of the accuracy for testing is 92.47%.  

Thus, using WaterCloud data we can even closer approach to solving the problem of drought within the California State Program – “20X2020 Water Conservation Plan” as accurately as possible without causing protests from the population and court proceeding.


May Issue 2023