Analysis of elevated temperature readings in dairy cattle using time-series analysis with subcutaneous temperature sensor data
| dc.contributor.advisor | Niine, Tarmo (advisor) | |
| dc.contributor.author | Kosonen, Casper | |
| dc.date.accessioned | 2024-06-11T11:35:22Z | |
| dc.date.available | 2024-06-11T11:35:22Z | |
| dc.date.defensed | 2024-06-12 | |
| dc.date.issued | 2024 | |
| dc.description | Final Thesis Curriculum in Veterinary Medicine | eng |
| dc.description.abstract | Monitoring temperature is an important part of monitoring the overall health status of cattle. Increasing herd sizes cause problems to farmers, because among other problems, restraining and measuring individual cows’ temperature is becoming extremely labor-intensive and difficult. Technologies like the subcutaneous temperature sensors are being studied to ease monitoring this aspect of welfare in cattle. Our hypothesis was that there would be increased volatility in the subcutaneous sensor data during periods of elevated environmental temperatures. Secondary objective was to analyze circadian rhythm dynamics between two farms with different housing systems during the whole observation period. In current study, 6 bull calves (Farm 1) and 11 calves (Farm 2) with implanted subcutaneous temperature sensors were observed throughout the summer. The observation period for the Farm 1 bull calves was 48 days from July until August in 2020, and for the Farm 2 calves 78 days from June until August in 2022. The bull calves lived in an outside group pen while the calves were housed in an uninsulated indoor barn. Environmental data was collected on temperature and humidity, and temperature humidity index (THI) was calculated at each farm during the whole observation period. Generalized autoregressive conditional heteroskedasticity (GARCH) models with one lagged value (1h) were used to assess the subcutaneous (SC) temperature volatility. The autoregressive conditional heteroscedasticity (ARCH) test was performed to assess the performance of GARCH model. The results showed that the humidity in the indoor housing system was more erratic and frequently high, while in the outside pen, humidity showed lower values and a daily seasonality. GARCH model results showed that alpha was very low and beta very high which indicates that in the context of calves’ body temperature, the persistence of volatility was high, but recency was low. The ARCH test showed that the accuracy of GARCH model was poor for some calves from each farm. The circadian rhythm of the bull calves had temperature peaks during the day and lowest points during the night, while the calves had peaks during the evening and lowest readings in the morning. | eng |
| dc.description.abstract | Temperatuuri jälgimine on veiste üldise tervisliku seisundi jälgimise oluline osa. Piimaveise karjade looma arvu suurenemine valmistab loomakasvatajatele probleeme, sest muude kitsaskohtade hulgas muutub äärmiselt tööjõumahukaks ja keeruliseks ka üksikute lehmade kehatemperatuuri jälgimine ja mõõtmine. Veiste heaolu jälgimise hõlbustamiseks uuritakse lootustandvaid tehnoloogiaid nagu subkutaansed temperatuuriandurid. Uuringu hüpoteesiks oli, et kõrgema keskkonnatemperatuuriga perioodidel suureneb subkutaansete temperatuuriandurite registreeritud andmete volatiilsus. Teiseks eesmärgiks oli analüüsida ööpäevarütmi dünaamikat kahe erineva pidamissüsteemiga farmi vahel kogu vaatlusperioodi jooksul. Uuringu läbiviimiseks jälgiti kogu suve perioodi kuut pullvasikat (Farm 1) ja 11 lehmvasikat (Farm 2), kellele olid implanteeritud nahaalused temperatuuriandurid. Pullvasikate vaatlusperiood kestis 2020. aasta juulist augustini (48 päeva) ja 2022. aasta juunist augustini 78 päeva lehmvasikate puhul. Pullvasikad elasid õueaedikus, lehmvasikad aga soojustamata laudas. Kogu vaatlusperioodi jooksul koguti mõlemas farmis keskkonnaandmeid temperatuuri ja õhuniiskuse kohta ning arvutati temperatuuriniiskuse indeks (THI). Subkutaanse (SC) temperatuuri volatiilsuse hindamiseks kasutati üldistatud autoregressiivseid tingimusliku heteroskedastilisuse (GARCH) mudeleid ühe ajaperioodi viitega (1 tund). GARCH mudeli toimivuse hindamiseks viidi läbi autoregressiivne tingimusliku heteroskedastilisuse (ARCH) test. Tulemused näitasid, et siseruumides oli õhuniiskus ebastabiilsem ja sageli kõrge, samas kui õues oli õhuniiskus madalam ja esines märgatavat igapäevast, hooajalist muutlikkust. GARCH mudeli tulemused näitasid, et mudeli alfa oli väga madal ja beeta väga kõrge, mis viitab sellele, et vasikate kehatemperatuuri kontekstis oli volatiilsus püsiv, kuid hiljutisuse efekt madal. ARCH test näitas, et GARCH mudeli täpsus oli mõlema farmi mõne vasika puhul ebatäpne. Pullvasikate ööpäevarütmi temperatuuri tipud olid päeval ja madalaimad öösiti, lehmvasikatel aga õhtuti ja madalaimad hommikul. | est |
| dc.identifier.uri | http://hdl.handle.net/10492/9393 | |
| dc.publisher | Eesti Maaülikool | |
| dc.rights | info:eu-repo/semantics/openAccess | |
| dc.subject | magistritööd | est |
| dc.subject | master thesis | eng |
| dc.subject | Calf | eng |
| dc.subject | THI | |
| dc.subject | circadian rhythm | eng |
| dc.subject | GARCH model | eng |
| dc.subject | volatility | eng |
| dc.title | Analysis of elevated temperature readings in dairy cattle using time-series analysis with subcutaneous temperature sensor data | eng |
| dc.title.alternative | Kuumasstressi analüüs piimaveistel, kasutades nahaaluse temperatuurianduri andmeid | est |
| dc.type | info:eu-repo/semantics/masterThesis | |
| rioxxterms.freetoread.startdate | 12.09.2024 |
Failid
Originaal pakett
1 - 1 1
Laen...
- Nimi:
- Kosonen_final thesis_2024_fulltext.pdf
- Suurus:
- 3.42 MB
- Formaat:
- Adobe Portable Document Format