{
"cells": [
{
"cell_type": "markdown",
"id": "f77e2e5e",
"metadata": {},
"source": [
"# Solar Decomposition Models\n",
"\n",
"Knowing the direct or beam normal irradiance (DNI) is useful for many solar and energy applications, e.g., calculating the yield of solar concetrating power systems or determining the irradiance on an inclined surface. However, often only global horizontal irradiance (GHI) is available, as it is much cheaper to measure than the individual components and most satelitte radiation models only calculate GHI. Many applications, therefore, require applying **solar radiation decomposition models** to obtain the direct beam irradiance. The majority of decomposition models work by applying an empirical correlations between GHI, diffuse horizontal irradiance (DHI) and DNI. In this section, we present the implementation in Python of several solar decomposition models. \n",
"\n",
"***\n",
"
Content by Javier Lopez Lorente and Adam R. Jensen
\n"
]
},
{
"cell_type": "markdown",
"id": "c349c5fe",
"metadata": {},
"source": [
"## The role of solar decomposition models\n",
"\n",
"Solar decomposition models are some of the most frequently used algorithms, and are often used during the first part of a larger analysis. As described earlier, decomposition models permit estimating the beam or direct normal irradiance (DNI) from GHI measurements, often using relationships between the clearness index ($k_t$) and the diffuse fraction of solar irradiance ($k_d$). The clearness index is defined as:\n",
"\n",
"$$ k_t = \\frac{GHI}{E_a \\cdot cos(\\theta_Z)}$$\n",
"\n",
"where $E_a$ is the extraterrestrial irradiance, and $\\theta_Z$ is the solar zenith angle. \n",
"\n",
"The diffuse fraction is defined as: \n",
"\n",
"$$ k_d = \\frac{DHI}{GHI} $$\n",
"\n",
"The role of decomposition models can be graphically illustrated in the diagram below:\n",
"\n",
"```{image} ../notebooks/graphics/decomposition_models.png\n",
":alt: decomposition_flowchart\n",
":class: bg-white mb-1\n",
":width: 500px\n",
":align: center\n",
"```\n"
]
},
{
"cell_type": "markdown",
"id": "e03414bf",
"metadata": {},
"source": [
"Multiple models have been proposed for the decomposition of global horizontal radiation/irradiance into its direct normal and diffuse horizontal components in the scientific literature. Below, the implementation in Python of some of these methods will be shown. The following methods will be modelled and compared:\n",
"- DISC model (Maxwell, 1987)\n",
"- DIRINT model (Pérez et al., 1992)\n",
"- DIRINDEX model (Pérez et al., 2002)\n",
"- Erbs model (Erbs et al., 1982)\n",
"\n",
"The four decomposition models, among some others, are implemented in the open-source library [pvlib python](https://pvlib-python.readthedocs.io/en/stable/api.html#dni-estimation-models).\n",
"\n",
"Before applying the models in practice, an example irradiance data set will be retrieved."
]
},
{
"cell_type": "markdown",
"id": "da7971ff",
"metadata": {},
"source": [
"## Data preparation\n",
"\n",
"To illustrate the implementation of each of these models, we will use real irradiance observations. The decomposition models will be used to calculate DNI and DIh from measured GHI, and the results will then be compared to measurements of DNI and DHI. The first step is getting and preparing the measurement data to be used."
]
},
{
"cell_type": "code",
"execution_count": 1,
"id": "c9acfda5",
"metadata": {},
"outputs": [],
"source": [
"# Import all the necessary libraries\n",
"import numpy as np\n",
"import pandas as pd\n",
"import matplotlib.pyplot as plt\n",
"import pvlib"
]
},
{
"cell_type": "markdown",
"id": "ec93ecdf",
"metadata": {},
"source": [
"### Retrieving weather data\n",
"\n",
"The irradiance data used is from the [BSRN](https://bsrn.awi.de/) Cabauw station operated by the Royal Netherlands Meteorological Institute (KNMI). The data is retrieved from the BSRN FTP server using the `get_bsrn` function from pvlib-python. The Cabauw station is located at latitude 51.9711° and longitude 4.9267°, at an elevation of 0 m AMSL. The reason for choosing the Cabauw station is that the data has already been extensively quality checked, hence in this example this step is skipped. The data used are 1-minute global, direct, and diffuse irradiance measurements between January and July 2021, which are resampled to hourly observations."
]
},
{
"cell_type": "code",
"execution_count": 2,
"id": "e3fbf2c9",
"metadata": {
"tags": [
"remove-cell"
]
},
"outputs": [],
"source": [
"# Load the BSRN credentials from the Secret environment variables\n",
"# This cell is hidden in the displayed notebook\n",
"import os\n",
"#bsrn_username = os.environ[\"BSRN_FTP_USERNAME\"]\n",
"#bsrn_password = os.environ[\"BSRN_FTP_PASSWORD\"]\n",
"bsrn_username = 'bsrnftp'\n",
"bsrn_password = 'bsrn1'\n",
"sodapro_username = 'assessingsolar@gmail.com'"
]
},
{
"cell_type": "code",
"execution_count": 3,
"id": "7f322efe",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
ghi
\n",
"
ghi_std
\n",
"
ghi_min
\n",
"
ghi_max
\n",
"
dni
\n",
"
dni_std
\n",
"
dni_min
\n",
"
dni_max
\n",
"
dhi
\n",
"
dhi_std
\n",
"
dhi_min
\n",
"
dhi_max
\n",
"
lwd
\n",
"
lwd_std
\n",
"
lwd_min
\n",
"
lwd_max
\n",
"
temp_air
\n",
"
relative_humidity
\n",
"
pressure
\n",
"
\n",
" \n",
" \n",
"
\n",
"
2021-01-01 00:00:00+00:00
\n",
"
-1.150000
\n",
"
0.076667
\n",
"
-1.366667
\n",
"
-1.033333
\n",
"
1.000000
\n",
"
0.008333
\n",
"
1.000000
\n",
"
1.133333
\n",
"
-1.000000
\n",
"
0.050000
\n",
"
-1.000000
\n",
"
-1.000000
\n",
"
242.133333
\n",
"
0.288333
\n",
"
241.666667
\n",
"
242.633333
\n",
"
-1.531667
\n",
"
99.360000
\n",
"
1006.000000
\n",
"
\n",
"
\n",
"
2021-01-01 01:00:00+00:00
\n",
"
-1.000000
\n",
"
0.078333
\n",
"
-1.000000
\n",
"
-1.000000
\n",
"
1.150000
\n",
"
0.005000
\n",
"
1.050000
\n",
"
1.550000
\n",
"
-1.000000
\n",
"
0.051667
\n",
"
-1.000000
\n",
"
-1.000000
\n",
"
243.666667
\n",
"
0.253333
\n",
"
243.216667
\n",
"
244.100000
\n",
"
-2.320000
\n",
"
98.990000
\n",
"
1006.000000
\n",
"
\n",
"
\n",
"
2021-01-01 02:00:00+00:00
\n",
"
-1.000000
\n",
"
0.081667
\n",
"
-1.016667
\n",
"
-0.900000
\n",
"
1.266667
\n",
"
0.016667
\n",
"
1.050000
\n",
"
1.516667
\n",
"
-0.600000
\n",
"
0.051667
\n",
"
-0.683333
\n",
"
-0.500000
\n",
"
265.950000
\n",
"
0.603333
\n",
"
264.983333
\n",
"
267.066667
\n",
"
-0.808333
\n",
"
98.183333
\n",
"
1006.066667
\n",
"
\n",
"
\n",
"
2021-01-01 03:00:00+00:00
\n",
"
-1.000000
\n",
"
0.075000
\n",
"
-1.000000
\n",
"
-1.000000
\n",
"
1.583333
\n",
"
0.006667
\n",
"
1.383333
\n",
"
1.616667
\n",
"
-0.533333
\n",
"
0.065000
\n",
"
-0.833333
\n",
"
-0.333333
\n",
"
271.800000
\n",
"
0.578333
\n",
"
270.800000
\n",
"
272.733333
\n",
"
-0.930000
\n",
"
98.736667
\n",
"
1006.216667
\n",
"
\n",
"
\n",
"
2021-01-01 04:00:00+00:00
\n",
"
-0.966667
\n",
"
0.076667
\n",
"
-1.000000
\n",
"
-0.850000
\n",
"
1.316667
\n",
"
0.005000
\n",
"
1.233333
\n",
"
1.500000
\n",
"
0.000000
\n",
"
0.061667
\n",
"
-0.350000
\n",
"
0.000000
\n",
"
291.933333
\n",
"
0.231667
\n",
"
291.500000
\n",
"
292.266667
\n",
"
-0.171667
\n",
"
97.960000
\n",
"
1006.050000
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" ghi ghi_std ghi_min ghi_max dni \\\n",
"2021-01-01 00:00:00+00:00 -1.150000 0.076667 -1.366667 -1.033333 1.000000 \n",
"2021-01-01 01:00:00+00:00 -1.000000 0.078333 -1.000000 -1.000000 1.150000 \n",
"2021-01-01 02:00:00+00:00 -1.000000 0.081667 -1.016667 -0.900000 1.266667 \n",
"2021-01-01 03:00:00+00:00 -1.000000 0.075000 -1.000000 -1.000000 1.583333 \n",
"2021-01-01 04:00:00+00:00 -0.966667 0.076667 -1.000000 -0.850000 1.316667 \n",
"\n",
" dni_std dni_min dni_max dhi dhi_std \\\n",
"2021-01-01 00:00:00+00:00 0.008333 1.000000 1.133333 -1.000000 0.050000 \n",
"2021-01-01 01:00:00+00:00 0.005000 1.050000 1.550000 -1.000000 0.051667 \n",
"2021-01-01 02:00:00+00:00 0.016667 1.050000 1.516667 -0.600000 0.051667 \n",
"2021-01-01 03:00:00+00:00 0.006667 1.383333 1.616667 -0.533333 0.065000 \n",
"2021-01-01 04:00:00+00:00 0.005000 1.233333 1.500000 0.000000 0.061667 \n",
"\n",
" dhi_min dhi_max lwd lwd_std \\\n",
"2021-01-01 00:00:00+00:00 -1.000000 -1.000000 242.133333 0.288333 \n",
"2021-01-01 01:00:00+00:00 -1.000000 -1.000000 243.666667 0.253333 \n",
"2021-01-01 02:00:00+00:00 -0.683333 -0.500000 265.950000 0.603333 \n",
"2021-01-01 03:00:00+00:00 -0.833333 -0.333333 271.800000 0.578333 \n",
"2021-01-01 04:00:00+00:00 -0.350000 0.000000 291.933333 0.231667 \n",
"\n",
" lwd_min lwd_max temp_air \\\n",
"2021-01-01 00:00:00+00:00 241.666667 242.633333 -1.531667 \n",
"2021-01-01 01:00:00+00:00 243.216667 244.100000 -2.320000 \n",
"2021-01-01 02:00:00+00:00 264.983333 267.066667 -0.808333 \n",
"2021-01-01 03:00:00+00:00 270.800000 272.733333 -0.930000 \n",
"2021-01-01 04:00:00+00:00 291.500000 292.266667 -0.171667 \n",
"\n",
" relative_humidity pressure \n",
"2021-01-01 00:00:00+00:00 99.360000 1006.000000 \n",
"2021-01-01 01:00:00+00:00 98.990000 1006.000000 \n",
"2021-01-01 02:00:00+00:00 98.183333 1006.066667 \n",
"2021-01-01 03:00:00+00:00 98.736667 1006.216667 \n",
"2021-01-01 04:00:00+00:00 97.960000 1006.050000 "
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Retrieving the measurement data\n",
"df, meta = pvlib.iotools.get_bsrn(\n",
" station='CAB',\n",
" start=pd.Timestamp('2021-01-01'),\n",
" end=pd.Timestamp('2021-07-31'),\n",
" username=bsrn_username,\n",
" password=bsrn_password)\n",
"\n",
"# Resample the 1-minute data to hourly\n",
"df = df.resample('1h').mean()\n",
"\n",
"# Display the first five rows of the dataset\n",
"df.head()"
]
},
{
"cell_type": "markdown",
"id": "5c028e80",
"metadata": {},
"source": [
"Besides GHI, DNI and DHI, this station has additional weather-related variables available (e.g., air temperature, humidity, and wind speed), as well as sensor-specific data. The details of the sensors and measurements available in each station can be found under the section \"Instrument history and meta data\" of each MIDC's station webpage. \n",
"\n",
"Let's list the variables available and then subset those that we will use in this example:"
]
},
{
"cell_type": "code",
"execution_count": 4,
"id": "079596f1",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['ghi', 'ghi_std', 'ghi_min', 'ghi_max', 'dni', 'dni_std', 'dni_min',\n",
" 'dni_max', 'dhi', 'dhi_std', 'dhi_min', 'dhi_max', 'lwd', 'lwd_std',\n",
" 'lwd_min', 'lwd_max', 'temp_air', 'relative_humidity', 'pressure'],\n",
" dtype='object')"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Visualize variables avilable\n",
"df.columns"
]
},
{
"cell_type": "markdown",
"id": "1b51487e",
"metadata": {},
"source": [
"It is often helpful to visualize the data in order to check that the measurements are within the expected ones. For example, a plot for a single day (18th July 2021):"
]
},
{
"cell_type": "code",
"execution_count": 5,
"id": "a1edfe57",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"
"
]
},
"metadata": {
"needs_background": "light"
},
"output_type": "display_data"
}
],
"source": [
"df.loc['2021-07-18', ['ghi','dhi','dni']].plot()\n",
"plt.title('Solar irradiance components, 18th July 2021')\n",
"plt.ylabel('Irradiance [W/m$^2$]')\n",
"plt.legend()"
]
},
{
"cell_type": "markdown",
"id": "a140dcc5",
"metadata": {},
"source": [
"For an overall representation of the data, it's useful to inspect the the key statistical metrics for the data:"
]
},
{
"cell_type": "code",
"execution_count": 6,
"id": "f5a3bd44",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
ghi
\n",
"
dni
\n",
"
dhi
\n",
"
temp_air
\n",
"
relative_humidity
\n",
"
pressure
\n",
"
\n",
" \n",
" \n",
"
\n",
"
count
\n",
"
5088.000000
\n",
"
5021.000000
\n",
"
5006.000000
\n",
"
5088.000000
\n",
"
5088.000000
\n",
"
5088.000000
\n",
"
\n",
"
\n",
"
mean
\n",
"
150.154482
\n",
"
129.608143
\n",
"
78.730762
\n",
"
9.613794
\n",
"
81.598316
\n",
"
1016.147409
\n",
"
\n",
"
\n",
"
std
\n",
"
223.501191
\n",
"
244.733398
\n",
"
109.759797
\n",
"
7.098631
\n",
"
14.189124
\n",
"
10.043831
\n",
"
\n",
"
\n",
"
min
\n",
"
-2.033333
\n",
"
-0.316667
\n",
"
-1.827586
\n",
"
-8.151667
\n",
"
32.516667
\n",
"
981.450000
\n",
"
\n",
"
\n",
"
25%
\n",
"
-1.000000
\n",
"
1.000000
\n",
"
-0.783333
\n",
"
4.133750
\n",
"
72.566667
\n",
"
1009.945833
\n",
"
\n",
"
\n",
"
50%
\n",
"
21.075000
\n",
"
1.316667
\n",
"
18.233333
\n",
"
8.776667
\n",
"
84.843333
\n",
"
1016.675000
\n",
"
\n",
"
\n",
"
75%
\n",
"
234.120833
\n",
"
117.466667
\n",
"
125.820833
\n",
"
15.312083
\n",
"
93.289167
\n",
"
1023.575000
\n",
"
\n",
"
\n",
"
max
\n",
"
926.033333
\n",
"
958.500000
\n",
"
568.216667
\n",
"
29.330000
\n",
"
100.000000
\n",
"
1042.000000
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" ghi dni dhi temp_air relative_humidity \\\n",
"count 5088.000000 5021.000000 5006.000000 5088.000000 5088.000000 \n",
"mean 150.154482 129.608143 78.730762 9.613794 81.598316 \n",
"std 223.501191 244.733398 109.759797 7.098631 14.189124 \n",
"min -2.033333 -0.316667 -1.827586 -8.151667 32.516667 \n",
"25% -1.000000 1.000000 -0.783333 4.133750 72.566667 \n",
"50% 21.075000 1.316667 18.233333 8.776667 84.843333 \n",
"75% 234.120833 117.466667 125.820833 15.312083 93.289167 \n",
"max 926.033333 958.500000 568.216667 29.330000 100.000000 \n",
"\n",
" pressure \n",
"count 5088.000000 \n",
"mean 1016.147409 \n",
"std 10.043831 \n",
"min 981.450000 \n",
"25% 1009.945833 \n",
"50% 1016.675000 \n",
"75% 1023.575000 \n",
"max 1042.000000 "
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df[['ghi','dni','dhi', 'temp_air', 'relative_humidity', 'pressure']].describe()"
]
},
{
"cell_type": "markdown",
"id": "60df7b39",
"metadata": {},
"source": [
"**Estimating other input variables:**\n",
"\n",
"The 3 solar decomposition models that we will implement in this section (i.e., DISC, DIRINT and Erbs models) have different inputs to be implemented. **Solar position data** is a common input for the 3 models. The DISC and DIRINT models accept as a possible input the **dew (wet-bulb) temperature** ($T_d$). This variable is not directly available from our weather observations; however, it can be easily estimated using a simple conversion that has relative humidity ($RH$) and the ambient temperature ($T_a$) as input. This simple approximation is given by the equation below (Lawrence, 2005):\n",
"\n",
" $$ T_{d} = T_a (^{\\circ} C) - \\frac{100-RH (\\%)}{5} $$\n",
" \n",
"In Python, this equation could be implemented as follows:"
]
},
{
"cell_type": "code",
"execution_count": 7,
"id": "51d5ea06",
"metadata": {},
"outputs": [],
"source": [
"df['temp_dew'] = df['temp_air']-((100-df['relative_humidity'])/5)"
]
},
{
"cell_type": "markdown",
"id": "2df5c477",
"metadata": {},
"source": [
" \n",
"\n",
"The solar position for the entire period can be calculated using the `pvlib.solarposition.get_solarposition` function which supports a number of different solar position algorithms. The default algorithm is an implementation of the NREL SPA algorithm."
]
},
{
"cell_type": "code",
"execution_count": 8,
"id": "7ab8defb",
"metadata": {},
"outputs": [
{
"data": {
"text/html": [
"
\n",
"\n",
"
\n",
" \n",
"
\n",
"
\n",
"
apparent_zenith
\n",
"
zenith
\n",
"
apparent_elevation
\n",
"
elevation
\n",
"
azimuth
\n",
"
equation_of_time
\n",
"
\n",
" \n",
" \n",
"
\n",
"
2021-01-01 00:00:00+00:00
\n",
"
149.693651
\n",
"
149.693651
\n",
"
-59.693651
\n",
"
-59.693651
\n",
"
21.451575
\n",
"
-3.441097
\n",
"
\n",
"
\n",
"
2021-01-01 01:00:00+00:00
\n",
"
144.585022
\n",
"
144.585022
\n",
"
-54.585022
\n",
"
-54.585022
\n",
"
45.257706
\n",
"
-3.460697
\n",
"
\n",
"
\n",
"
2021-01-01 02:00:00+00:00
\n",
"
137.059810
\n",
"
137.059810
\n",
"
-47.059810
\n",
"
-47.059810
\n",
"
63.685127
\n",
"
-3.480288
\n",
"
\n",
"
\n",
"
2021-01-01 03:00:00+00:00
\n",
"
128.334219
\n",
"
128.334219
\n",
"
-38.334219
\n",
"
-38.334219
\n",
"
78.296833
\n",
"
-3.499870
\n",
"
\n",
"
\n",
"
2021-01-01 04:00:00+00:00
\n",
"
119.152361
\n",
"
119.152361
\n",
"
-29.152361
\n",
"
-29.152361
\n",
"
90.721936
\n",
"
-3.519442
\n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" apparent_zenith zenith apparent_elevation \\\n",
"2021-01-01 00:00:00+00:00 149.693651 149.693651 -59.693651 \n",
"2021-01-01 01:00:00+00:00 144.585022 144.585022 -54.585022 \n",
"2021-01-01 02:00:00+00:00 137.059810 137.059810 -47.059810 \n",
"2021-01-01 03:00:00+00:00 128.334219 128.334219 -38.334219 \n",
"2021-01-01 04:00:00+00:00 119.152361 119.152361 -29.152361 \n",
"\n",
" elevation azimuth equation_of_time \n",
"2021-01-01 00:00:00+00:00 -59.693651 21.451575 -3.441097 \n",
"2021-01-01 01:00:00+00:00 -54.585022 45.257706 -3.460697 \n",
"2021-01-01 02:00:00+00:00 -47.059810 63.685127 -3.480288 \n",
"2021-01-01 03:00:00+00:00 -38.334219 78.296833 -3.499870 \n",
"2021-01-01 04:00:00+00:00 -29.152361 90.721936 -3.519442 "
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# Define a pvlib location object for the site\n",
"location = pvlib.location.Location(latitude=51.9711, longitude=4.9267, altitude=0, name='Cabauw')\n",
"\n",
"# Calculate the solar position for each index in the datarame for the specific location\n",
"# The 30 minute shift is applied in order to calculate the solar position at the middle of the hour!\n",
"solpos = location.get_solarposition(df.index+pd.Timedelta(minutes=30))\n",
"solpos.index = solpos.index - pd.Timedelta(minutes=30)\n",
"\n",
"# Let's have a look to the output\n",
"solpos.head()"
]
},
{
"cell_type": "markdown",
"id": "30588eae",
"metadata": {},
"source": [
"## DISC model\n",
"\n",
"The DISC model (Maxwell, 1987) requires the following inputs: GHI, solar zenith angle, and the atmospheric pressure, which is used to estimate the absolute (pressure-corrected) airmass."
]
},
{
"cell_type": "code",
"execution_count": 9,
"id": "159e2691",
"metadata": {},
"outputs": [],
"source": [
"# DISC method estimated with absolute airmass as input\n",
"disc = pvlib.irradiance.disc(\n",
" ghi=df['ghi'],\n",
" solar_zenith=solpos['apparent_zenith'],\n",
" datetime_or_doy=df.index,\n",
" pressure=df['pressure']*100) # Site pressure in Pascal"
]
},
{
"cell_type": "markdown",
"id": "15df0a83",
"metadata": {},
"source": [
"The functoin returns the DNI, the clearness index ($k_t$) and the airmass estimation:"
]
},
{
"cell_type": "code",
"execution_count": 10,
"id": "e7cec5ac",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['dni', 'kt', 'airmass'], dtype='object')"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"disc.columns"
]
},
{
"cell_type": "markdown",
"id": "6c575ce0",
"metadata": {},
"source": [
"## DIRINT model\n",
"\n",
"An implementatin of the DIRINT model (Pérez et al., 1992) is also available in pvlib-python. The model requires the following inputs: GHI, solar zenith angle, atmospheric pressure, and dew (wet-bulb) temperature."
]
},
{
"cell_type": "code",
"execution_count": 11,
"id": "d0689dda",
"metadata": {},
"outputs": [],
"source": [
"# Estimation of the DIRINT model\n",
"dirint = pvlib.irradiance.dirint(\n",
" ghi=df['ghi'],\n",
" solar_zenith=solpos['apparent_zenith'],\n",
" times=df.index,\n",
" pressure=df['pressure']*100, # atmospheric pressure in Pascal \n",
" temp_dew=df['temp_dew']) # Dew temperature in Degree Celsius"
]
},
{
"cell_type": "markdown",
"id": "b8131d76",
"metadata": {},
"source": [
"## DIRINDEX model\n",
"The DIRINDEX model (Pérez et al., 2002) was found to be one of the two best decomposition models in an extensive study by Gueymard and Ruiz-Arias (2016). The model is an enhancement of the DIRINT model by considering clearsky data."
]
},
{
"cell_type": "code",
"execution_count": 12,
"id": "518222ec",
"metadata": {},
"outputs": [],
"source": [
"# Retrieve clear-sky irradiance from CAMS McClear\n",
"clearsky, meta = pvlib.iotools.get_cams(\n",
" latitude=location.latitude,\n",
" longitude=location.longitude,\n",
" altitude=location.altitude,\n",
" start=df.index[0],\n",
" end=df.index[-1],\n",
" email=sodapro_username)\n",
"\n",
"# Estimation of the DIRINDEX model\n",
"dirindex = pvlib.irradiance.dirindex(\n",
" ghi=df['ghi'],\n",
" ghi_clearsky=clearsky['ghi_clear'],\n",
" dni_clearsky=clearsky['dni_clear'],\n",
" zenith=solpos['zenith'],\n",
" times=df.index,\n",
" pressure=df['pressure']*100, # atmospheric pressure in Pascal \n",
" temp_dew=df['temp_dew']) # Dew temperature in Degree Celsius"
]
},
{
"cell_type": "markdown",
"id": "a67adee7",
"metadata": {},
"source": [
"## Erbs model\n",
"\n",
"The Erbs model (Erbs et al., 1982) is a simple and very popular model that only requires GHI and the solar zenith angle. This model is also available in pvlib and the implementation is equally straight-forward:"
]
},
{
"cell_type": "code",
"execution_count": 13,
"id": "3a9d9856",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"Index(['dni', 'dhi', 'kt'], dtype='object')"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"erbs = pvlib.irradiance.erbs(\n",
" ghi=df['ghi'],\n",
" zenith=solpos['apparent_zenith'],\n",
" datetime_or_doy=df.index)\n",
"\n",
"erbs.columns"
]
},
{
"cell_type": "markdown",
"id": "5d5f6aef",
"metadata": {},
"source": [
"## Comparison and evaluation of decomposition models\n",
"\n",
"Once the 4 models are estimated, we can evaluate the results of the models against the actual DNI measurements available. For example, below we visualize the results of the models a single day and a week in July:\n"
]
},
{
"cell_type": "code",
"execution_count": 14,
"id": "91ffaf95",
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
""
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"