Accurate and timely nowcasting is vital in safeguarding public property and ensuring people's safety, particularly in regions with infrequent but potentially hazardous rainfall events like the United Arab Emirates (UAE). The UAE experiences rare rainfall occurrences that can cause flooding, making it crucial to address these weather challenges. In this study, we compared the predictability of two widely used ensemble-based nowcasting systems, STEPS and LINDA, to assess their performance during various precipitation events in the UAE. Our ROC analysis showed that both systems have a commendable detection rate of over 0.70, a low false alarm rate of less than 0.2, and a substantial area under the curve exceeding 0.75%. Additionally, both algorithms demonstrated their ability to produce reliable nowcasts for up to two hours, using a 5 mm/h threshold. These findings highlight the potential of these models in mitigating the impacts of rare rainfall events, such as flooding, and emphasize the importance of investing in advanced nowcasting technologies for improved weather prediction and public safety.