Injin Koyon Na'ura don Ƙididdigar Evapotranspiration na Tunani tare da Ƙarancin Bayanan Yanayi

Teburin Abubuwan Ciki

1 Gabatarwa

Ƙididdigar evapotranspiration na tunani (ETo) tana da muhimmanci ga shirin ban ruwa da sarrafa albarkatun ruwa, musamman a yankuna masu fama da fari kamar Morocco. Ma'aunin FAO-56 Penman-Monteith, ko da yake yana da daidaito, yana buƙatar ɗimbin bayanan yanayi da suka haɗa da zafin jiki, zafi, hasken rana, da saurin iska, wanda hakan ya sa ba zai yiwu a yankuna masu ƙarancin kayan aikin firikwensin ba.

Ma'auni na gargajiya na zahiri irin su Hargreaves-Samani, Romanenko, da Jensen-Haise suna ba da hanyoyin sauƙaƙa amma suna fama da bambancin aiki a cikin yanayin yanayi daban-daban. Wannan binciken yana magance waɗannan iyakoki ta hanyar bincika samfurorin koyon injin waɗanda zasu iya samun daidaitaccen ƙididdigar ETo tare da mafi ƙarancin sigogi.

2 Hanyar Aiki

2.1 Tattara Bayanai da Gyara su

An tattara bayanan yanayi daga tashoshi da yawa a yankin Meknes, ciki har da ma'aunin zafin jiki, zafi, hasken rana, da saurin iska. Gyaran bayanai ya haɗa da sarrafa ƙimar da suka ɓace, daidaitawa, da daidaita lokaci a cikin tashoshi daban-daban.

2.2 Samfurorin Koyon Injin

An aiwatar da samfurorin koyon injin guda uku kuma an kwatanta su:

• XGBoost: Tsarin haɓakawa wanda aka sani da babban aiki da inganci
• Injin Tallafawa Vector (SVM): Yana da tasiri ga ayyukan koma bayan tare da ƙarancin bayanai
• Random Forest (RF): Hanyar haɗakarwa mai ƙarfi ga wuce gona da iri

2.3 Tsarin Gwaji

An aiwatar da matakan tabbatarwa guda biyu:

• Scenario 1: Rarraba duk bayanan da ake da su ba da gangan ba
• Scenario 2: Horarwa a kan tasha ɗaya, tabbatarwa a kan wani (tabbatar da rarrabuwar kawuna ta sarari)

3 Sakamako da Tattaunawa

3.1 Kwatancen Aiki

Duk samfurorin koyon injin sun fi ma'auni na gargajiya na zahiri a cikin duka matakan tabbatarwa. XGBoost ya nuna mafi girman daidaito tare da ƙimar R² wanda ya wuce 0.92, sannan Random Forest da SVM suka biyo baya.

3.2 Binciken Muhimmancin Fasali

Zafin jiki da hasken rana sun fito a matsayin mafi mahimmancin fasali a cikin dukkan samfurorin. Binciken ya nuna cewa tare da waɗannan sigogi biyu kawai, samfurorin koyon injin zasu iya samun kashi 85-90% na aikin da aka samu tare da cikakkun saitin sigogi.

4 Aiwatar da Fasaha

4.1 Tsarin Lissafi

Ma'aunin FAO-56 Penman-Monteith na daidaitaccen ma'auni ne:

$$ET_0 = \frac{0.408\Delta(R_n - G) + \gamma\frac{900}{T + 273}u_2(e_s - e_a)}{\Delta + \gamma(1 + 0.34u_2)}$$

Inda $\Delta$ shine gangaren lanƙwasa matsi tururi, $R_n$ shine radiation net, $G$ shine magudanar zafi na ƙasa, $\gamma$ shine akai psychrometric, $T$ shine zafin jiki, $u_2$ shine saurin iska, $e_s$ shine matsin tururi cikakke, kuma $e_a$ shine ainihin matsin tururi.

4.2 Aiwatar da Lambar

import xgboost as xgb
from sklearn.ensemble import RandomForestRegressor
from sklearn.svm import SVR
import numpy as np

class EToEstimator:
    def __init__(self, model_type='xgb'):
        if model_type == 'xgb':
            self.model = xgb.XGBRegressor(
                max_depth=6,
                learning_rate=0.1,
                n_estimators=100,
                objective='reg:squarederror'
            )
        elif model_type == 'rf':
            self.model = RandomForestRegressor(
                n_estimators=100,
                max_depth=10,
                random_state=42
            )
        elif model_type == 'svm':
            self.model = SVR(kernel='rbf', C=1.0, epsilon=0.1)
    
    def train(self, X_train, y_train):
        self.model.fit(X_train, y_train)
    
    def predict(self, X_test):
        return self.model.predict(X_test)

# Zaɓin fasali: zafin jiki da hasken rana kawai
features = ['temp_max', 'temp_min', 'solar_rad']
target = 'ETo_FAO56'

5 Aikace-aikacen Gaba

Binciken ya nuna babban yuwuwar aiwatarwa a aikace a fagage da yawa:

• Tsarin Ban Ruwa na Wayo: Haɗa kai tare da masu sarrafa ban ruwa na IoT don sarrafa ruwa na ainihi
• Daidaituwa Canjin Yanayi: Ingantaccen tsara albarkatun ruwa a yankuna masu fama da fari
• Fasahar Noma: Haɓaka aikace-aikacen wayar hannu don ƙananan manoma
• Manufar Ruwa: Tallafin yanke shawara na tushen bayanai don raba ruwa da farashi

Hanyoyin bincike na gaba sun haɗa da koyon canja wuri a cikin yankuna daban-daban na yanayi, haɗa kai tare da bayanan tauraron dan adam, da haɓaka hanyoyin ƙididdiga na gefe don yankuna masu nisa.

6 Nassoshi

1. Allen, R. G., Pereira, L. S., Raes, D., & Smith, M. (1998). Evapotranspiration amfanin gona: Jagororin don ƙididdigar bukatun ruwan amfanin gona. Takardar ban ruwa da magudanar ruwa ta FAO 56.
2. Landeras, G., Ortiz-Barredo, A., & López, J. J. (2008). Kwatanta samfurorin hanyar sadarwar jijiyoyi na wucin gadi da ma'auni na zahiri da rabin zahiri don ƙididdigar evapotranspiration na tunai na yau da kullun a cikin Ƙasar Basque. Gudanar da Ruwan Noma, 95(5), 553-565.
3. Maestre-Valero, J. F., Martínez-Alvarez, V., & González-Real, M. M. (2013). Kimanta SVM da ELM don ƙididdigar evapotranspiration na tunai na yau da kullun a yankuna masu bushewa. Kwamfuta da Kayan aikin Lantarki a Noma, 89, 100-106.
4. López-Urrea, R., Martín de Santa Olalla, F., Fabeiro, C., & Moratalla, A. (2006). Gwada ma'auni na evapotranspiration ta amfani da abin lura na lysimeter a cikin yanayi mai bushewa. Gudanar da Ruwan Noma, 85(1-2), 15-26.
5. Goodfellow, I., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S., ... & Bengio, Y. (2014). Gidajen adawa masu samarwa. Ci gaban tsarin bayanai na jijiyoyi, 27.

7 Binciken Kwararre