Abstract: An intelligent orchard monitoring system integrating multi-source sensing, multi-mode communications, and edge-cloud collaborative processing has been proposed to achieve precise real-time monitoring of orchard environments．Modular design has combined multiple sensor types including soil sensors（temperature, humidity, pH, electrical conductivity）, meteorological sensors（air temperature, humidity, wind speed, wind direction, atmospheric pressure）, illuminance, and carbon dioxide concentration to enable comprehensive acquisition of environmental parameters．A hybrid power supply scheme using a 20 W photovoltaic panel and a 6600 mAh lithium battery, coupled with multi-level low power strategy, supports stable operation for more than 7 continuous days during overcast or rainy weather．Field testing data from an 8 month trial in 3.33 hm² apple orchard in Yantai City, Shandong Province, showed that overall measurement accuracy within ±2%（root mean square error of soil temperature reduced from ±1.8 °C to ± 0.6 °C）, communication success rate of 98.7%（15.3% improvement over single LoRa）, and end-to-end data processing latency ≤ 2.3 seconds．In practical applications, system enabled precision irrigation with 35% water savings, provided pest and disease early warnings seven days in advance（reducing pesticide use by 40%）, and yielded a payback period of 2.1 years．