9:25 AM - 9:45 AM
[ACG48-02] Observation of vertical distribution of water temperature and cross-sectional distribution of river velocity using an electric reel
★Invited Papers
Keywords:Coastal environmental monitoring, UAV with electric winch, electric reel, in situ observation
Ship surveys or instruments installed at points have been used to measurewater temperatures and other things in coastal areas, including brackish water areas. In recent years, areal observations with UAV have provided the waves and turbidity on the ocean surface from the images by UAVs. We have been developing a fixed-point vertical observation system by UAV-mounted winch that descends or ascends CTD for observations in areas where ships cannot navigate (like shallow water, lakes and harbors where ships cannot leave due to freezing). However, the vertical distribution of water temperature and other parameters differs between ascending and descending due to the effect of the sensor's response speed, which necessitates the consideration of the sensor's movement speed. Vertical measurement of water quality and cross-sectional measurement of river current velocity by a small boat equipped with a velocimeter require towing the instrument with by a cord, etc. Therefore, it is desirable to measure and record at regular time intervals at a constant and low towing speed as much as possible. In recent years, inexpensive electric reels for fishing and lines that can withstand high loads have become available. The utilization of real makes the observation with towing more ideally, and, consequently, facilitates the study of movement speed relative to sensor response. In order to use the winch for observations in the ocean and to understand the response speed of a water temperature and salinometer to the winding speed of a winch-mounted UAV, we conducted vertical observations of water temperature and cross-sectional observations of river current velocity using an electric reel, as well as laboratory experiments.
The water temperature was measured at the deepest site in the Obuchi Lake by CTD. Cross-sectional observations of river velocities were conducted around the river mouth of the Obuchi River by a riverboat equipped with an ADCP. We used the SonTek CastAway-CTD (hereafter CA-CTD) that is not constrained by the height of the lower part of the UAV winch and the payload. The reproducibility of the vertical distribution of water temperature was examined using water temperature and water depth measured every 0.2 seconds along ascending paths from the bottom by seven winding speeds as the mode displayed on the electric reel changed from 1 to 7. The response speed of both water temperature sensors was estimated by alternately acclimating and moving the CA-CTD and RINKO-Profiler (hereafter referred to as RINKO) to buckets filled with hot and cold water. Cross-sectional observation using an ADCP was conducted by tying a thin string to piles at both ends of the river, which were guide ropes (the width of the river was about 40 m), attaching a pulley with a riverboat equipped with an ADCP to the guide rope, and towing the pulley manually from the right bank and by an electric reel from the left bank.
The electric reel did not wind smoothly below a certain speed, and the vibration of the winding speed induced the oscillations in the vertical distribution of the water temperature. The electric reel in this study required a winding speed of 10 cm s-1 or faster. The vertical distribution of water temperature obtained with the CA-CTD wound up at high speed is shifted upward relative to the vertical distribution observed with the CTD wound up at low speed because the water temperature sensor is unable to accurately track the actual water temperature. The lag correlations of these vertical profiles indicated that the vertical profiles of water temperatures by CA-CTD at windup velocities of less than 20 cm s-1 (lag of 0 cm). The half-lives of the initial temperature difference between the sensor and the site, meaning the response time of CA-CTD and RINKO, were estimated to be approximately 0.4 and 0.1 seconds, respectively, from the experiment. The towing a riverboat with an electric reel (except mode1) enable to observe the vertical distribution of current velocity with less oscillation of vessel speed and tilt (improved accuracy) and lower speed (higher resolution) than with manual towing. In addition to these cross-sectional velocity observations, a riverboat-mounted ADCP was driven by Wiral Lite, which can automatically move back and forth over the motion camera cable like a gondola. Water pressure was measrured for a long period at both the lake and the river. The time series of water height was corrected from the pressure data by using the sea surface height monitored with GPS (+ ichimill) on the float.
The water temperature was measured at the deepest site in the Obuchi Lake by CTD. Cross-sectional observations of river velocities were conducted around the river mouth of the Obuchi River by a riverboat equipped with an ADCP. We used the SonTek CastAway-CTD (hereafter CA-CTD) that is not constrained by the height of the lower part of the UAV winch and the payload. The reproducibility of the vertical distribution of water temperature was examined using water temperature and water depth measured every 0.2 seconds along ascending paths from the bottom by seven winding speeds as the mode displayed on the electric reel changed from 1 to 7. The response speed of both water temperature sensors was estimated by alternately acclimating and moving the CA-CTD and RINKO-Profiler (hereafter referred to as RINKO) to buckets filled with hot and cold water. Cross-sectional observation using an ADCP was conducted by tying a thin string to piles at both ends of the river, which were guide ropes (the width of the river was about 40 m), attaching a pulley with a riverboat equipped with an ADCP to the guide rope, and towing the pulley manually from the right bank and by an electric reel from the left bank.
The electric reel did not wind smoothly below a certain speed, and the vibration of the winding speed induced the oscillations in the vertical distribution of the water temperature. The electric reel in this study required a winding speed of 10 cm s-1 or faster. The vertical distribution of water temperature obtained with the CA-CTD wound up at high speed is shifted upward relative to the vertical distribution observed with the CTD wound up at low speed because the water temperature sensor is unable to accurately track the actual water temperature. The lag correlations of these vertical profiles indicated that the vertical profiles of water temperatures by CA-CTD at windup velocities of less than 20 cm s-1 (lag of 0 cm). The half-lives of the initial temperature difference between the sensor and the site, meaning the response time of CA-CTD and RINKO, were estimated to be approximately 0.4 and 0.1 seconds, respectively, from the experiment. The towing a riverboat with an electric reel (except mode1) enable to observe the vertical distribution of current velocity with less oscillation of vessel speed and tilt (improved accuracy) and lower speed (higher resolution) than with manual towing. In addition to these cross-sectional velocity observations, a riverboat-mounted ADCP was driven by Wiral Lite, which can automatically move back and forth over the motion camera cable like a gondola. Water pressure was measrured for a long period at both the lake and the river. The time series of water height was corrected from the pressure data by using the sea surface height monitored with GPS (+ ichimill) on the float.