Using artificial intelligence in the mapping of Quaternary units

With the advancement of knowledge and technology, this has led to liberalization and public access to medium and large-scale spatiotemporal data in a variety of spectra. What resulted in more than three decades of continuous monitoring of the earth's crust from above the atmosphere in different orbits by multisensory satellites is now set on a public platform and are accessible for research, government and commercial use. Unlimited access to all pre-existing location-based data, the ability to schedule and process very quickly in the server's cloud environment with pre-prepared samples and the ability to load data in addition to what is available in the server, opened a new perspective on redefining the use of multisensory and server satellite data in recent years.  The process and quality of producing server-based thematic maps is in absolute dependence on the input data, both pre-existing and generated data, in the course of the geological study process.Choosing a suitable platform for processing information, spectral data and radar of satellite images and even using aerial photographs appropriate to the desired scale has a significant and undeniable role in preparing and increasing the accuracy of the initial map. It is natural to achieve the maximum and possible final accuracy in the verification process by repeating step by step field observations and sampling and, of course, combining it several times with laboratory results.Rehabilitation and optimization of age, lithological and geochemical data of each separated rock unit, in addition to using reasoned and significant information published from the integration of laboratory data obtained from sampling performed in multiple stages of field control will come.Field verification based on the initial map, the possibility of access and targeted observation of one or more times separated from each rock unit resulting from the process of data processing and satellite image, not only in the quadrilateral area of ​​the map, but even beyond the study area cover. The nature of layered and multiple processing of spatial information and its online integration with pre-existing data, both in the area and in the surrounding areas, are among the factors that determine the final uncertainty coefficient in proportion to the amount and accuracy of processed information and data set. Doing this process not only improves the accuracy and uncertainty of server-based second-generation geological maps from qualitative coefficient (speculation) to numerical coefficient, but also determines the error coefficient of each unit relative to the surrounding unit. Clearly, however, at the end of this stage, not only is it possible to edit and correct (manually) point-by-point map correlations with satellite and aerial imagery with higher accuracy, which is necessary. It should be noted again that there is no doubt that the degree of accuracy and the final uncertainty coefficient depend on the accuracy and clarity of input data, quantity and quality of field surveys and at the end of case editing and correction of the map in proportion to the output scale of the target map. With this introduction, the aim of this project is to formulate and develop new methodologies to prepare geological thematic maps based on pre-existing satellite data and cloud computing process on the public servers such as GEE combined with field and laboratory controls. Second generation geological-thematic maps are: rapidity of studies, integration of legends and geological units, significant reduction in the use of instrumental analysis, simultaneous saving of the process of updating information production with precision and global standards. The 1: 100000 scale geological map of the Kavire-e Lut is the first of eleven geological sheets prepared with this new protocol. The map includes a collection of volcanic, pyroclastic and sedimentary rock deposits from Triassic to Quaternary in the western part of Lut block within the scope of the work active seismic fault performance of Nayband.