You should provide more details. The choice of the method of preparing river water samples for spectrometric measurements depends on the following conditions and the measuring and auxiliary equipment available. Should the object of analysis be water including suspension or do you also need separate results for the filtrate? Which elements are you interested in and what detection and quantification limits for each element do you need to achieve? What AAS techniques do you have at your disposal? Only FAAS, or also GFAAS, HGAAS? Which spectrometer model do you have (manufacturer, type)? Are you an experienced AAS spectrometer user or rather a beginner?

In general, the most useful information, guidelines and methodological recommendations can be found in information and training materials available on the websites of spectrometer manufacturers and in various publications of official, professional and scientific institutions dealing with environmental research, in particular surface and ground waters. If you are a beginner in the AAS technique, a training stay in a specialized laboratory, equipped with a spectrometer technically similar to your spectrometer will be the best for you.

The procedure is described in the article whose link is: Article A Comparison of Reproducibility of Inductively Coupled Spect...

In addition to Zbigniew Jońca 's answer, why do you need to digest samples of water? The matrix is already liquid, whereas digestion is needed to solubilize solid material (soil, plants, food, etc) into a form that can be pumped into the analytical instrument.

The typical method for water samples is acidification with approximately 1% HNO3 or similar to keep the dissolved minerals in solution until analysis (and prevent biological metabolism from bacteria/phytoplankton etc into different forms).

There are many sampling protocols out there. Try the US EPA documents such as: https://www.epa.gov/sites/default/files/2017-02/documents/sampling_guidance_for_unknown_contaminants_in_drinking_water_02152017_final.pdf