# Identification of Optimal Starting Time Instance to Forecast Net Blotch Density in Spring Barley with Meteorological Data in Finland

^{1}

^{2}

^{*}

## Abstract

**:**

## 1. Introduction

^{−1}yield-loss in spring barley in the long term in the Nordic countries. An assessment in 2015 showed that 40% of fields in South Tigray, in Ethiopia, had net blotch and 60% of them had its relative, spot blotch [5]. This can decrease the barley crop by 10–20% of the annual average yield [6,7], but yield losses as high as 40% have been reported [8].

## 2. Materials and Methods

#### 2.1. Data

- Category 1 (very low net blotch density, maximum net blotch severity value of 0.5%);
- Category 2 (net blotch appears in the selected observation fields in these years, severity value of 0.6−5%).

- Atmospheric pressure (hPa);
- Relative humidity (RH %);
- Temperature (°C);
- Dew point temperature (°C).

#### 2.2. General Structure of Data Analysis

_{min}) and the maximum value (x

_{max}) were found from the Category 1 data (low net blotch density). The Category 2 (high net blotch density) data were normalised with their corresponding minimum and maximum values. The beginning of the growing season varies according to the year and observation field because of the varying climate conditions and the geological position. This has been considered in the analysis by selecting the starting point of each data set at the beginning of the growing season instead of a fixed calendar date, as explained below in the Section 2.3.

#### 2.3. Starting Date of Growing Season, Automatic Calculation

#### 2.4. Feature Generation

_{sn}value (see Equations (2) and (3) below).

- If the relative humidity is >87%, then the leaf is humid → LWD = 1;
- If the relative humidity is >70%–<87% and increasing >3% per 30 min, then the leaf is humid → LWD = 1;
- If the relative humidity is >70%–<87% and decreasing >2% per 30 min, then the leaf is dry → LWD = 0;
- If the relative humidity <70%, then the leaf is dry → LWD = 0.

#### 2.5. Metrics

_{sn}, the distance between two vectors a and b, Category 1 and 2, respectively, with their mean values µ

_{a}and µ

_{b}and standard deviations δ

_{a}and δ

_{b}are computed according to Equation (2) [38]:

_{sn}for the identification of the optimal starting time instance for net blotch prediction proceeds in sliding windows from the beginning of the estimated growing season and the following 50 days, step by step in data windows of 14 days for every generated feature n as a sum of the calculated D

_{sn}daily values:

_{C}

_{1}and M

_{C}

_{2}are feature matrices generated from scalar observations of weather variables related to the years with data of Categories 1 and 2 (see Table 1), ${\overline{x}}_{nj}$ is the average of the data of the years in question and s

_{nj}is the standard deviation of the same data. In Figure 4, the behaviour of the D

_{sn}index is illustrated. There, Feature number 1 would be ranked as a more plausible candidate than Feature number 2 by comparing their calculated D

_{sn}values in classification of the two groups (o and x). According to Figure 4 and Equation (2) with the same notation of statistical quantities, the resulting value of D

_{sn}for Feature 1 would be much higher than for Feature 2 in this case.

## 3. Results and Discussion

_{sn}values in the fourteen-day time window are presented for each time step. The daily average, minimum and maximum of the studied variables were tested and the D

_{sn}values were calculated accordingly to every analysed feature subset (Equation (3)).

_{sn}values for the average, minimum and maximum outdoor temperature are presented. The highest D

_{sn}value, 52.5, is achieved on day 40 from the beginning of the growing season with the daily average outdoor temperature. The D

_{sn}value in the case of the average outdoor temperature remains relatively high between days 37 and 40, but then the value falls rapidly to 28.2 on day 41. On the other hand, the D

_{sn}value of the maximum outdoor temperature achieves its highest value, 43.3, on the same day while the highest D

_{sn}value of the minimum outdoor temperature of 29.3 is achieved on day 29. The D

_{sn}value of the average outdoor temperature higher than the D

_{sn}values of the minimum and the maximum outdoor temperature during the studied period. This indicates that the information content of the outdoor temperature related to the appearance of barley net blotch is the highest during the two weeks starting on days 36–40 from the beginning of the growing season.

_{sn}values of the average, minimum and maximum daily relative humidity are presented. The highest D

_{sn}value, namely 57, is achieved on day 22 from the beginning of the growing season with the minimum daily relative humidity. The average of the daily relative humidity reaches its highest D

_{sn}value of 45.9 at almost the same time, namely on day 25. The maximum daily relative humidity exhibits the highest D

_{sn}values in the time window 42–49 days from the beginning of the growing season. As can be seen in Figure 5, the D

_{sn}values are relatively high between days 17 and 30 with all three variables, but the maximum daily relative humidity does not achieve its highest value of 49.5 until day 49.

_{sn}values of the daily average, minimum and maximum dew point temperature are presented. The highest D

_{sn}value, 48.9, is achieved for this weather variable on day 39 from the beginning of the growing season (day 1) with the daily average dew point temperature. In the case of the daily maximum dew point temperature, the D

_{sn}value (42.1) peaks in the same time window, whereas the highest D

_{sn}value related to the minimum dew point temperature (39.6) is achieved on day 29. The classification potential of the dew point temperature to separate the two data sets related to different levels of net blotch risk increases during the growing season until day 40, which can be seen in the ascending trend of all three series in Figure 7.

_{sn}values for the daily average, minimum and maximum atmospheric pressure are shown. The highest D

_{sn}value, 74.4 is achieved here on day 14 when applying the measured minimum values of the atmospheric pressure. Another peak appears on day 18 and corresponds to a D

_{sn}value of 71.4. Furthermore, the highest D

_{sn}values of the maximum (49.1; day 13) and average atmospheric pressure (68.4; day 14) peak almost in the same starting time instance. During days 21–28, the D

_{sn}values of all three statistical quantities for atmospheric pressure are relatively low. All the D

_{sn}values increase slightly after day 30, but are still considerably lower than during days 11–19.

_{sn}values for leaf wetness duration (LWD) are presented. The highest D

_{sn}value, 34.5 is achieved when the calculation is started on day 22 from the beginning of the growing season. Relatively high D

_{sn}values also exist between days 27 and 32. The time window when the maximum D

_{sn}values are achieved differs from the peaks presented in Figure 4, Figure 5, Figure 6 and Figure 7. Here, the maximum D

_{sn}value of LWD is at a lower level than the D

_{sn}values of the other analysed weather variables.

_{sn}values for the 715 best-ranked features calculated in data windows of 14 days at each starting time instance (day). On each daily boxplot, the central mark indicates the median of the 715 calculated D

_{sn}values of the related features. The bottom and top edges of the boxplot indicate the 25th and 75th percentiles, respectively. The whiskers extend to the most extreme data points and the individual high D

_{sn}values are plotted with ‘o’ markers.

_{sn}value, namely 285.8, is achieved on day 41 from the beginning of the growing season with the generated feature prototype number 115 (Appendix A), which is structured here as the combination of three weather variables:

_{sn}value 260.9), 40 (D

_{sn}value 256.2) and 42 (D

_{sn}value 276.9). The highest statistical median for the D

_{sn}value of the plotted features appears on day 18, namely 92.9. It can also be concluded from Figure 10 in comparison to the values of single weather variables (Figure 5, Figure 6, Figure 7, Figure 8 and Figure 9) that the classification potential generally increases with the applied features. The highest D

_{sn}values of analysed variables and of the best generated feature are presented in Table 2.

_{sn}values) of the best-ranked feature (Equation (2)) is compared at the starting time instances of days 41 and 18 from the beginning of the growing season. The daily D

_{sn}values are presented with standard deviations. The markers ‘x’ and ‘o’ are the mean values of Category 1 (low net blotch density) and 2 (high net blotch density) data, respectively, and the whiskers describe standard deviation, namely the interval with a confidence level of 68%. In Figure 11, the D

_{sn}values with Category 1 data are generally higher than those of Category 2 for every monitored day. Statistically, the categories differ from each other during nine days out of 14 with a confidence level of 68%, as can be seen from Figure 11.

_{sn}values of Categories 1 and 2 are similar on five days out of 14 and the whiskers overlap in every case, namely with a confidence level of 68%. Thus, statistically the D

_{sn}values for the best feature (Equation (2)) are similar in both data sets, leading to poor classification potential.

## 4. Conclusions

## Author Contributions

## Funding

## Data Availability Statement

## Conflicts of Interest

## Appendix A

features(1) = x − y; |

features(2) = x − z; |

features(3) = y − z; |

features(4) = (x − y) × y; |

features(5) = (y − x) × z; |

features(6) = (z − x) × z; |

features(7) = (y − z) × z; |

features(8) = (z − y) × x; |

features(9) = (x − z) × y; |

features(10) = ln(x); |

features(11) = ln(y); |

features(12) = ln(z); |

features(13) = x × y; |

features(14) = x × z; |

features(15) = x × y × z; |

features(16) = y × z; |

features(17) = ln(x) − ln(y); |

features(18) = ln(x) − ln(z); |

features(19) = ln(y) − ln(z); |

features(20) = ln(x) − ln(y) × ln(z); |

features(21) = ln(y) − ln(x) × ln(y); |

features(22) = ln(z) − ln(x) × ln(z); |

features(23) = ln(y) − ln(z) × ln(z); |

features(24) = ln(z) − ln(y) × ln(x); |

features(25) = ln(x)/ln(y); |

features(26) = ln(x) × ln(y); |

features(27) = ln(x) × ln(z); |

features(28) = ln(x) × ln(y) × ln(z); |

features(29) = ln(y) × ln(z); |

features(30) = sqrt(x); |

features(31) = sqrt(y); |

features(32) = sqrt(z); |

features(33) = sqrt(x) − sqrt(y); |

features(34) = sqrt(x) − sqrt(z); |

features(35) = sqrt(y) − sqrt(z); |

features(36) = sqrt(ln(x)); |

features(37) = sqrt(ln(y)); |

features(38) = sqrt(ln(z)); |

features(39) = sqrt(x)/y; |

features(40) = x/z; |

features(41) = y/z; |

features(42) = (x × y)/z; |

features(43) = (x × z)/y; |

features(44) = (y × z)/x; |

features(45) = sqrt(x)/sqrt(y); |

features(46) = sqrt(x)/z; |

features(47) = (y/x)^2; |

features(48) = (sqrt(x) × y)/z; |

features(49) = (sqrt(x) × z)/y; |

features(50) = (y × z)/sqrt(x); |

features(51) = x^2; |

features(52) = y^2; |

features(53) = z^2; |

features(54) = x^2 − y^2; |

features(55) = x^2 − z^2; |

features(56) = x; |

features(57) = y; |

features(58) = z; |

features(59) = x + y + z; |

features(60) = x + y − z; |

features(61) = ln(x) + ln(y) + ln(z); |

features(62) = sqrt(y) + sqrt(z) + sqrt(x); |

features(63) = (x − y)/x; |

features(64) = (x/y)^3; |

features(65) = (y^(0.7) − 1)/(0.7); |

features(66) = (y − z)/y; %(y − z)/z, 23.12.2011 |

features(67) = (z − y)/x; |

features(68) = (y^(−1) − 1)/(−1); |

features(69) = x + y; |

features(70) = x + z; |

features(71) = y + z; |

features(72) = (x + y)/y; |

features(73) = (y + x)/z; |

features(74) = (y^(0.5) − 1)/(0.5); |

features(75) = (z^(2.5) − 1)/(2.5); |

features(76) = (z + y)/x; |

features(77) = (y^(1.5) − 1)/(1.5); |

features(78) = (x + z)/x; |

features(79) = (y^(−2) − 1)/(−2); |

features(80) = (x + z)/y; |

features(81) = ln(x) + ln(y); |

features(82) = ln(x) + ln(z); |

features(83) = ln(y) + ln(z); |

features(84) = (ln(x) + ln(y)) × ln(z); |

features(85) = (ln(y) + ln(x)) × ln(y); |

features(86) = (ln(z) + ln(x)) × ln(z); |

features(87) = (ln(y) + ln(z)) × ln(z); |

features(88) = (ln(z) + ln(y)) × ln(x); |

features(89) = (ln(x) + ln(z)) × ln(y); |

features(90) = sqrt(x) + sqrt(y); |

features(91) = sqrt(x) + sqrt(z); |

features(92) = sqrt(y) + sqrt(z); |

features(93) = (x + y) × y; |

features(94) = (y + x) × z; |

features(95) = (z + x) × z; |

features(96) = (y + z) × z; |

features(97) = (z + y) × x; |

features(98) = (x + z) × y; |

features(99) = (x + z) × x; |

features(100) = (x − y) × x; |

features(101) = x + (y × y); |

features(102) = y + (x × z); |

features(103) = z + (x × z); |

features(104) = y + (z × z); |

features(105) = z + (y × x); |

features(106) = x + (z × y); |

features(107) = x + (z × x); |

features(108) = x − (y × x); |

features(109) = y^2 − z^2; |

features(110) = x^2 × y^2; |

features(111) = (x − y) × z; |

features(112) = (x + y) × z; |

features(113) = (x/y) × z; |

features(114) = (x/y) + z; |

features(115) = ln(x)/ln(y) × ln(z); |

## Appendix B

The weather data used has been downloaded from the fmi open database: https://www.ilmatieteenlaitos.fi/havaintojen-lataus#!/ (accessed on 2 October 2022) |

Mynämäki: until 2011, the FMI weather station “Turku airport” and 2012–2017 the FMI weather station “Kaarina, Yltöinen”. |

Jokioinen: the FMI weather station “Jokioinen”. |

Seinäjoki: the FMI weather station “Seinäjoki, Pelmaa”. |

Siikajoki: the FMI weather station “Siikajoki, Revonlahti”. |

## References

- Ruusunen, O.; Jalli, M.; Jauhiainen, L.; Ruusunen, M.; Leiviskä, K. Data Analysis in Moving Windows for Optimizing Barley Net Blotch Prediction. J. Adv. Agric. Technol.
**2020**, 7, 154–196. [Google Scholar] [CrossRef] - FAO. FAOSTAT. 2020. Available online: http://www.fao.org/faostat/en/ (accessed on 21 January 2022).
- Jalli, M.; Laitinen, P.; Latvala, S. The emergence of cereal fungal diseases and the incidence of leaf spot diseases in Finland. Agric. Food Sci.
**2011**, 20, 62–73. [Google Scholar] [CrossRef] - Jalli, M.; Kaseva, J.; Andersson, B.; Ficke, A.; Nistrup-Jørgensen, L.; Ronis, A.; Kaukoranta, T.; Ørum, J.-E.; Djurle, A. Yield increases due to fungicide control of leaf blotch diseases in wheat and barley as a basis for IPM decision-making in the Nordic-Baltic region. Eur. J. Plant Pathol.
**2020**, 158, 315–333. [Google Scholar] [CrossRef] - Teferi, T.A.; Wubshet, M.L.; Aregawi, T.B. Occurrence and intensity of net and spot blotch of barley in South Tigray, Ethiopia. Glob. Sci. Res. J.
**2015**, 3, 113–123. [Google Scholar] - Agriculture Victoria Net blotches of barley. 2020. Available online: https://agriculture.vic.gov.au/biosecurity/plant-diseases/grain-pulses-and-cereal-diseases/net-blotches-of-barley (accessed on 1 September 2022).
- El Yousfi, B.; Ezzahiri, B. Net Blotch on semi-arid regions of Morocco II—Yield and yield-loss modelling. Field Crops Res.
**2002**, 73, 81–93. [Google Scholar] [CrossRef] - Jayasena, K.W.; Van Burgel, C.A.; Tanaka, K.; Majewski, J.; Loughman, R. Yield reduction in barley in relation to spot-type net blotch. Australas. Plant Pathol.
**2007**, 36, 429–433. [Google Scholar] [CrossRef] - Turkington, T.K.; Tekauz, A.; Xi, K.; Kutcher, H.R. Foliar diseases of barley: Don’t rely on a single strategy from the disease management toolbox. Prairies Soils Crops J.
**2011**, 4, 142–150. [Google Scholar] - Aktar, W.; Sengupta, D.; Chowdhury, A. Impact of pesticides use in agriculture: Their benefits and hazards. Interdiscip. Toxicol.
**2009**, 2, 1–12. [Google Scholar] [CrossRef] [Green Version] - European Union. Directive 2009/128/EC of the European Parliament and the Council of 21 October 2009: Establishing a Framework for Community Action to Achieve the Sustainable use of Pesticides. Off. J. Eur. Union
**2009**, 309, 71–86. [Google Scholar] - European Commission. Green Deal: Pioneering Proposals to Restore Europe’s Nature by 2050 and Halve Pesticide Use by 2030. 2022. Available online: https://ec.europa.eu/commission/presscorner/detail/en/ip_22_3746 (accessed on 1 September 2022).
- Charaya, M.U.; Upadhyay, A.; Bhati, H.P.; Kumar, A. Plant disease forecasting: Past practices to emerging technologies. In Plant Disease: Management Strategies; Nehra, S., Ed.; Agrobios Research: Rajasthan, India, 2021; pp. 1–30. [Google Scholar]
- Fenu, G.; Malloci, F.M. Forecasting Plant and Crop Disease: An Explorative Study on Current Algorithms. Big Data Cogn. Comput.
**2021**, 5, 2. [Google Scholar] [CrossRef] - Gent, D.H.; Mahaffee, W.F.; McRoberts, N.; Pfender, W.F. The Use and Role of Predictive Systems in Disease Management. Annu. Rev. Phytopathol.
**2013**, 51, 267–289. [Google Scholar] [CrossRef] [PubMed] [Green Version] - Jørgensen, L.N.; Matzen, N.; Ficke, A.; Andersson, B.; Jalli, M.; Ronis, A.; Nielsen, G.C.; Erlund, P.; Djurle, A. Using risk models for control of leaf blotch diseases in barley minimises fungicide use—Experiences from the Nordic and Baltic countries. Acta Agric. Scand. Sect. B Soil Plant Sci.
**2021**, 71, 247–260. [Google Scholar] [CrossRef] - Secher, B.J.M.; Jørgensen, L.N.; Murali, N.S.; Boll, P.S. Field validation of a decision support system for the control of pests and diseases in cereals in denmark. Pestic. Sci.
**1995**, 45, 195–199. [Google Scholar] [CrossRef] - Henriksen, K.E.; Jørgensen, L.N.; Nielsen, G.C. PC-plant protection—A tool to reduce fungicide input in winter wheat, winter barley and spring barley in Denmark. In Proceedings of the Brighton Crop Protection Conference—Pest and Diseases, Brighton, UK, 13–16 November 2000; pp. 835–840. [Google Scholar]
- Bligaard, J.; Jørgensen, L.N.; Axelsen, J.; Hansen, J.G.; Ørum, J.E.; Baby, S.; Nielsen, G.C. Udvikling af Nye Risikomodeller for Septoria (Zymoseptoria tritici) i Vinterhvede; Miljø-og Fødevareministeriet, Miljøstyrelsen. Bekæmpelsesmiddelforskning: Odense, Denmark, 2017; p. 168. ISBN 978-87-93529-68-7. [Google Scholar]
- WisuEnnuste. 2022. Available online: https://www.minunmaatilani.fi/ohjelmistot-ja-palvelut/viljelysuunnitteluohjelmat/wisuennuste-kasvinsuojelun-tasmalliseen-ajoittamiseen (accessed on 3 February 2022).
- El Jarroudi, M.; Kouadio, A.L.; El Jarroudi, M.; Junk, J.; Bock, C.; Diouf, A.A.; Delfosse, P. Improving fungal disease forecasts in winter wheat: A critical role of intra-day variations of meteorological conditions in the development of Septoria leaf blotch. Field Crops Res.
**2017**, 213, 12–20. [Google Scholar] [CrossRef] - Fernando, W.G.D.; Oghenekaro, A.O.; Tucker, J.R.; Badea, A. Building on a foundation: Advances in epidemiology, resistance breeding, and forecasting research for reducing the impact of fusarium head blight in wheat and barley. Can. J. Plant Pathol.
**2021**, 43, 495–526. [Google Scholar] [CrossRef] - Landschoot, S.; Waegeman, W.; Audenaert, K.; Van Damme, P.; Vandepitte, J.; De Baets, B.; Haesaert, G. A field-specific web tool for the prediction of Fusarium head blight and deoxynivalenol content in Belgium. Comput. Electron. Agric.
**2013**, 93, 140–148. [Google Scholar] [CrossRef] - Musa, T.; Hecker, A.; Vogelgsang, S.; Forrer, H.R. Forecasting of Fusarium head blight and deoxynivalenol content in winter wheat with FusaProg. EPPO Bull.
**2007**, 37, 283–289. [Google Scholar] [CrossRef] - Shah, D.A.; Paul, P.A.; De Wolf, E.D.; Madden, L.V. Predicting plant disease epidemics from functionally-represented weather series. Phil. Trans. R. Soc. B
**2019**, 374, 20180273. [Google Scholar] [CrossRef] [Green Version] - Shah, D.A.; De Wolf, E.D.; Paul, P.A.; Madden, L.V. Functional data analysis of weather variables linked to Fusarium head blight epidemics in the United States. Phytopathology
**2019**, 109, 96–110. [Google Scholar] [CrossRef] [Green Version] - Shah, D.A.; De Wolf, E.D.; Paul, P.A.; Madden, L.V. Predicting Fusarium head blight epidemics with boosted regression tree. Phytopathology
**2014**, 104, 702–714. [Google Scholar] [CrossRef] [Green Version] - Shah, D.A.; Molineros, J.E.; Paul, P.A.; Willyerd, K.T.; Madden, L.V.; De Wolf, E.D. Predicting Fusarium head blight epidemics with weather-driven pre- and post-anthesis logistic regression model. Phytopathology
**2013**, 103, 906–919. [Google Scholar] [CrossRef] [Green Version] - Ruusunen, O.; Jalli, M.; Jauhiainen, L.; Ruusunen, M.; Leiviskä, K. Advanced Data Analysis as a Tool for Net Blotch Density Estimation in Spring Barley. Agriculture
**2020**, 10, 179. [Google Scholar] [CrossRef] - Saari, E.E.; Prescott, M. A scale for appraising the foliar intensity of wheat diseases. Plant Dis. Rep.
**1975**, 59, 377–379. [Google Scholar] - Blum, A.L.; Langley, P. Selection of relevant features and examples in machine learning. Artif. Intell.
**1997**, 97, 245–271. [Google Scholar] [CrossRef] - Dash, M.; Liu, H. Feature selection for classification. Intell. Data Anal.
**1997**, 1, 131–156. [Google Scholar] [CrossRef] - García-Torres, M.; Gómez-Vela, F.; Melián-Batista, B.; Moreno-Vega, J.M. High-dimensional feature selection via feature grouping: A Variable Neighborhood Search approach. Inf. Sci.
**2016**, 326, 102–118. [Google Scholar] [CrossRef] - Pérez-Rodríguez, J.; Arroyo-Peña, A.G.; García-Pedrajas, N. Simultaneous instance and feature selection and weighting using evolutionary computation: Proposal and study. Appl. Soft Comput.
**2015**, 37, 416–443. [Google Scholar] [CrossRef] - Uncu, Ö.; Türkşen, I.B. A novel feature selection approach: Combining feature wrappers and filters. Inf. Sci.
**2007**, 177, 449–466. [Google Scholar] [CrossRef] - Ruusunen, M. Signal Correlations in Biomass Combustion—An Information Theoretic Analysis. Acta Univ. Ouluensis Ser. C
**2013**, 459, 1–120. [Google Scholar] - Kruit, R.J.W.; van Pul, W.A.J.; Jacobs, A.F.G.; Heusinkveld, B.G. Comparison between four methods to estimate leaf wetness caused by dew on grassland. In Proceedings of the 26th Conference on Agricultural and Forest Meteorology, Session 10.1, Vancouver, BC, Canada, 23–26 August 2004. [Google Scholar]
- Pomeroy, S.; Tamayo, P.; Gaasenbeek, M.; Sturla, L.M.; Angelo, M.; McLaughlin, M.E.; Kim, J.Y.H.; Goumnerovak, L.C.; Blackk, P.M.; Lau, C.; et al. Prediction of central nervous system embryonal tumour outcome based on gene expression. Nature
**2002**, 415, 436–442. [Google Scholar] [CrossRef] [PubMed]

**Figure 1.**Analysis procedure for meteorological data to identify time instance of optimal information content concerning forecasting the severity of net blotch occurrence.

**Figure 3.**Difference between the sowing date and the beginning of the growing season, presented by year (running number on x-axis) and observation fields (names).

**Figure 4.**An illustrated example of the usage of D

_{sn}for identifying classification properties of features in the case of two data groups (o and x). Classification between the data groups would be successful with Feature 1 in this case.

**Figure 5.**Variation of the classification potential (D

_{sn}values) with daily minimum, maximum and average outdoor temperatures. Day 1 on the x-axis is the first day of the growing season.

**Figure 6.**Variation of the classification potential (D

_{sn}values) when using the daily minimum, maximum and average values of relative humidity. Day 1 is the beginning of the growing season.

**Figure 7.**The variation of the classification potential (D

_{sn}values) for net blotch risk levels when applying the daily minimum, maximum and average values of the dew point temperature. Day 1 is the beginning of the growing season.

**Figure 8.**The variation of the classification potential (D

_{sn}values) of the daily minimum, maximum and average values of atmospheric pressure. Day 1 is the beginning of the growing season.

**Figure 9.**Variation of the classification potential (D

_{sn}values) with daily calculated leaf wetness duration. Day 1 on the x-axis is the first day of the growing season.

**Figure 10.**Box plots of the classification potential (D

_{sn}values) when using the 115 feature prototypes with 715 different variable combinations at each starting day in time windows of 14 days. The highest individual D

_{sn}values are plotted with ‘o’ markers.

**Figure 11.**D

_{sn}values and their standard deviations with Category 1 and 2 data applying the best feature (Equation (3)) and starting time on day 41 from the beginning of the growing season (14-day data window).

**Figure 12.**The mean D

_{sn}values and their standard deviations with Categories 1 and 2 data sets applying the best feature (Equation (3)) and starting the analysis on day 18 from the beginning of the growing season.

Location of Test Fields | Mynämäki N = 6,732,402.033 E = 218,702.907 | Jokioinen N = 6,746,822.331 E = 308,359.757 | Seinäjoki N = 6,986,750.229 E = 271,138.563 | Siikajoki N = 7,174,584.799 E = 408,818.353 | Years in Total |
---|---|---|---|---|---|

Years of observations Category 1 | 2011 | 2013 | 2011 | 2010 | 4 |

Years of observations Category 2 | 2013, 2014, 2016 | 2014, 2015 | 2016 | 2012, 2014, 2015 | 9 |

Variable | Highest D_{sn} Value | Time of the Best D_{sn} Value | Related Figure | |
---|---|---|---|---|

Daily outdoor temperature | Avg | 52.5 | 40 | 5 |

Min | 29.3 | 29 | ||

Max | 43.3 | 40 | ||

Relative humidity | Avg | 45.9 | 25 | 6 |

Min | 57 | 22 | ||

Max | 49.5 | 49 | ||

Dew point temperature | Avg | 48.9 | 39 | 7 |

Min | 39.6 | 29 | ||

Max | 42.1 | 39 | ||

Atmospheric pressure | Avg | 68.4 | 14 | 8 |

Min | 74.4 | 14 | ||

Max | 49.1 | 13 | ||

LWD | 34.5 | 22 | 9 | |

Feature with the highest D_{sn} value | 285.8 | 41 | 10 |

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. |

© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Ruusunen, O.; Jalli, M.; Jauhiainen, L.; Ruusunen, M.; Leiviskä, K.
Identification of Optimal Starting Time Instance to Forecast Net Blotch Density in Spring Barley with Meteorological Data in Finland. *Agriculture* **2022**, *12*, 1939.
https://doi.org/10.3390/agriculture12111939

**AMA Style**

Ruusunen O, Jalli M, Jauhiainen L, Ruusunen M, Leiviskä K.
Identification of Optimal Starting Time Instance to Forecast Net Blotch Density in Spring Barley with Meteorological Data in Finland. *Agriculture*. 2022; 12(11):1939.
https://doi.org/10.3390/agriculture12111939

**Chicago/Turabian Style**

Ruusunen, Outi, Marja Jalli, Lauri Jauhiainen, Mika Ruusunen, and Kauko Leiviskä.
2022. "Identification of Optimal Starting Time Instance to Forecast Net Blotch Density in Spring Barley with Meteorological Data in Finland" *Agriculture* 12, no. 11: 1939.
https://doi.org/10.3390/agriculture12111939