The invention discloses a deep bittern exploration evaluation method, which comprises the following steps of: 1, constructing a multi-scale
earth surface detection
system and establishing a three-dimensional geologic model; 2, implementing
deep drilling by adopting an optimal and fast drilling process, and synchronously carrying out
well logging and
well logging collaborative interpretation; the method comprises the following steps of 1, obtaining dynamic reservoir parameters through a single-hole multi-layer brine extraction test and a high-temperature and high-pressure in-situ test, 4, carrying out resource quantity intelligent
estimation based on a three-dimensional geological modeling and
machine learning
algorithm, and 5, outputting a three-dimensional visual exploration
decision scheme. According to the method, the electromagnetic-seismic joint inversion method is adopted, the constraint equation is established, the effective improvement of the longitudinal resolution of the stratum is achieved, the constructed multi-scale
earth surface detection
system can efficiently identify the reservoir at high precision, the precision reaches 0.5 m, the drilling efficiency is improved by adopting the optimal and fast drilling process, meanwhile, the accident rate is reduced, it is ensured that the conditions of the brine extraction (
discharge) test are met, and the method is suitable for large-scale production. And the exploration technology and the actual effect of the deep liquid
boron ore are practically improved.