Page 589 - Trends in Science and Technology fo Sustainable Living

550 Fakultas Sains dan Teknologi
Universitas Terbuka (2023)

Data and information from weather stations around the
salt ponds still need to be improved. BMKG’s weather monitoring
equipment is usually installed around airports, ports, plantations,
and rice fields. The need for fast, accurate, and up-to-date data
and information related to changes in the microclimate in the salt
pond environment, such as temperature, humidity, air pressure, and
rainfall, is also essential during salt production (Amin et al., 2021).
Therefore, it is necessary to install weather monitoring devices in the
salt pond environment to obtain accurate and real-time weather
information.
The availability of weather information through the tools
installed around the salt ponds can also explain salt farmers’
habits and experience in predicting the weather for salt production
purposes (Kuncoro et al., 2021; Widyanto et al., 2022). Even if we rely
only on classical predictions, the weather is not erratic, and the
season has changed its rhythm lately; of course, it will confuse salt
farmers in managing their ponds (Trikobery et al., 2017; Widyanto
et al., 2022). Due to this disorder, the habits and experiences of salt
farmers based on local wisdom could be more helpful. Therefore, the
presence of information technology coupled with local wisdom will
help determine the right time to start and harvest salt commodities
at a time like this so farmers can still optimize their production.

CONCLUSION

Data on rainfall and the number of rainy days is only feasible
to predict the fluctuations of salt yields in Cirebon. However, they
cannot accurately indicate the number of harvests. Predictions in this
study still use rainfall reanalysis data. This data must be validated
with observed rainfall data at specific salt pond locations. Thus, a
weather monitoring device is needed at specific salt pond locations
to determine a more accurate forecast of sea salt production to
measure rainfall over a long period. Further predictions can also be
developed by applying deep learning/machine learning methods,
especially when the correlation between variables tends to be